Kjell Gunnar Robbersmyr

• Nguyen, Thien Phuoc; Eid, Mahmoud Sayed Mahmoud; Huynh, van Khang & Robbersmyr, Kjell Gunnar (2026). Physics-Guided Cycle-Consistency Training for Stator Winding Fault Diagnosis in Synchronous Generators without Experimental Fault Data. IEEE Transactions on Transportation Electrification. ISSN 2577-4212. doi: 10.1109/TTE.2026.3727756. Fulltekst i vitenarkiv Vis sammendrag

  Stator inter-turn short circuit faults in shipboard synchronous generators (SGs) can rapidly escalate from minor insulation degradation into catastrophic failures, threatening the operational continuity and safety of electrified vessels. Therefore, early fault severity estimation of SGs is essential to prevent extensive downtime and ensure the resilience of shipboard power systems. However, the conventional data-driven approaches typically require labeled faulty data, which are rarely obtainable in practical maritime operations. To address this limitation, this article proposes a machine learning framework that enables early fault severity estimation without the need for experimental fault samples for training. The proposed method couples two neural networks through a physics-guided cycle-consistency (PGCC) training strategy. This approach facilitates both high-fidelity synthetic data generation and domain adaptation from analytical simulations to experimental measurements. The framework is validated on an in-house 7.5 kW test bench, demonstrating that the proposed method achieves high-accuracy prediction of unseen experimental fault severities. Ablation studies confirm the contribution of each network component and the PGCC training strategy. Furthermore, a comparative study demonstrates that the proposed method outperforms the tested domain adaptation and generative adversarial networks-based methods, particularly under limited training data.

• Dalal, Anurag; Hagen, Daniel; Robbersmyr, Kjell Gunnar & Knausgård, Kristian Muri (2026). Fast 3D Foundation Model Initialized Gaussian Splatting. 2026 6th International Conference on Electrical, Computer and Energy Technologies (ICECET). doi: 10.1109/icecet65726.2026.11633076. Fulltekst i vitenarkiv Vis sammendrag

  This paper introduces a fast method for high-quality 3D Gaussian Splatting (3DGS) reconstruction without traditional Structure-from-Motion (SfM). The proposed approach leverages 3D Foundation Models (3DFMs) for camera pose and pointcloud initialization, then jointly optimizes both camera poses and Gaussian primitives using a depth-guided loss function. This enables fast convergence even from rough initialization with as few as 50-60 input views. To further improve reconstruction quality in sparse-view scenarios, an MLP-based pose refinement module is introduced alongside depth-guided supervision from the foundation model. Extensive experiments on Mip-NeRF 360, Tanks and Temples, and RobustNeRF demonstrate that the proposed method achieves competitive reconstruction quality (23.61 dB PSNR, 0.19 LPIPS) while reducing training time to approximately three minutes per scene. The proposed method produces ready-to-use 3DGS models at a fraction of the time required by existing pipelines, making it suitable for near realtime applications in robotics, VR, and autonomous navigation.

• Nguyen, Hoang Nam Du; Huynh, Khang & Robbersmyr, Kjell Gunnar (2025). Domain Knowledge-Informed Few-Shot Fault Diagnosis of Multiple Low-Severity Faults in PMSGs Under Imbalanced Loads. IEEE transactions on industrial electronics (1982. Print). ISSN 0278-0046. s. 1–12. doi: 10.1109/TIE.2025.3629348. Fulltekst i vitenarkiv Vis sammendrag

  Electrical experts can readily diagnose high-severity faults in permanent magnet synchronous generators (PMSGs) using a few samples of harmonic order data in the time-frequency domain, leveraging their prior knowledge of interturn short circuits, local demagnetization, and mixed faults. However, traditional harmonic analysis methods are often hindered by low signal amplitude, noise, and overlap with various operating conditions, which present significant challenges for intelligent fault detection and identification models. Moreover, the high similarity between healthy and faulty signatures, combined with the scarcity of faulty samples across diverse operating scenarios, leads to high misclassification rates in conventional FDI models. To address these challenges, this article proposes an explainable few-shot learning architecture that integrates symmetrical components and harmonic analysis to classify incipient faults using only a small number of samples under multiple operating conditions. A convolutional neural network–based attention mechanism is employed to extract physical features of PMSGs in both normal and abnormal states. The proposed model achieves 89% accuracy with 20 shots, outperforming the pretrained ResNet50 (43%) on a 4320 sample PMSG dataset. The precision, recall, and F1-score metrics are used for validating the predictions. Additionally, by incorporating physical domain knowledge, the model can generate interpretable predictions with visual explanations of its inference process, aligning with the transparency and accountability principles mandated by the EU AI Act for high-risk systems in the energy sector.

• Nguyen, Thien Phuoc; Ha, Duy Hung; Huynh, Khang & Robbersmyr, Kjell Gunnar (2025). Harmonic-Based Analytical Model for Detecting Stator-Winding Short Circuit and Eccentricity Faults in Wound-Field Synchronous Generators. IEEE Access. 13, s. 195539–195551. doi: 10.1109/ACCESS.2025.3633506. Fulltekst i vitenarkiv Vis sammendrag

  Conventional analytical models for fault detection in salient-pole synchronous generators often ignore the influence of higher-order force harmonics, limiting their diagnosis accuracy. This paper proposes a harmonic-based analytical model, integrating the Maxwell stress tensor and airgap flux density, and accounting for both radial and tangential components of force under various fault conditions to investigate the impact of inter-turn short circuit (ITSC) and rotor eccentricity (ECC) faults on the force density and vibration characteristics in wound-field synchronous generators (WFSMs). The effects of ITSC and ECC faults are analyzed through spatiotemporal harmonic analysis, showing significant shifts in the harmonic spectrum with additional spatial harmonics at higher frequencies. The proposed analytical model is numerically verified through finite element analysis and experimentally validated on an in-house testbed of healthy and faulty WFSMs. The results prove that the proposed model effectively detects fault-related variations, particularly through higher spatial and temporal harmonic components whereas existing models primarily focus on the fundamental frequency components.

• Dalal, Anurag; Hagen, Daniel; Robbersmyr, Kjell Gunnar & Knausgård, Kristian Muri (2025). ScatteringSplatting: Capturing Minute Details Using Deep Learning-Aided Gaussian Splatting. International Conference on Control, Mechatronics and Automation. ISSN 2837-5114. s. 547–554. doi: 10.1109/iccma67641.2025.11369605. Fulltekst i vitenarkiv Vis sammendrag

  3D reconstruction in scattering media, such as underwater or foggy environments, presents significant challenges as well as substantial opportunities. Recent advances in novel view synthesis and accurate 3D reconstruction have enabled applications across diverse domains—from ecosystem monitoring and virtual observatories to underwater inspection and exploration using autonomous underwater vehicles and remotely operated vehicles. While Neural Radiance Fields excel at capturing environments with complex lighting effects, their long training times can limit practical applications. In contrast, 3D Gaussian Splatting offers a faster and more explainable alternative for 3D reconstruction and novel view synthesis. The proposed ScatteringSplatting method integrates volumetric rendering with 3D Gaussian Splatting to efficiently reconstruct underwater and foggy scenes. It explicitly separates the scattering media effects, handled via a multi-layer perceptron, from the 3D object geometries processed through Gaussian Splatting, and uses a unique loss function and effective culling of Gaussians to remove unwanted floaters. The primary advantage of our approach becomes evident when generating detailed views, particularly in rendering precise textures and small objects, such as fish. Our method achieves average PSNR, SSIM, and LPIPS scores of 28.91, 0.905, and 0.149, respectively, on the SeaThru-NeRF dataset.

• Dalal, Anurag; Hagen, Daniel; Robbersmyr, Kjell Gunnar & Knausgård, Kristian Muri (2025). Evaluating the effect of image quantity on Gaussian Splatting: A statistical perspective. Computer Vision and Image Understanding. ISSN 1077-3142. 263. doi: 10.1016/j.cviu.2025.104575. Fulltekst i vitenarkiv Vis sammendrag

  3D reconstruction is now a key capability in computer vision. With the advancements in NeRFs and Gaussian Splatting, there is an increasing need on properly capturing data to feed these algorithms and use them in real world scenarios. Most publicly available datasets that can be used for Gaussian Splatting are not suitable to do proper statistical analysis on reducing the number of cameras or the effect of uniformly placed cameras versus randomly placed cameras. The number of cameras in the scene significantly affects the accuracy and resolution of the final 3D reconstruction. Thus, designing a proper data capture system with a certain number of cameras is crucial for 3D reconstruction. In this paper UnrealGaussianStat dataset is introduced, and a statistical analysis is performed on decreasing viewpoints have on Gaussian splatting. It is found that when the number of cameras is increased after 100 the train and test metrics saturates, and does not have significant impact on the reconstruction quality.

• Eid, Mahmoud Sayed Mahmoud; Ali, Ahmed Mohamed Salem; Huynh, Khang & Robbersmyr, Kjell Gunnar (2025). Robust Self-augmented Open-circuit Fault Diagnosis of Three-level Inverters for EV Powertrains. IEEE Transactions on Transportation Electrification. ISSN 2577-4212. 11(4), s. 10250–10261. doi: 10.1109/TTE.2025.3554526. Fulltekst i vitenarkiv Vis sammendrag

  With the rapid advancement of data collection systems, data-driven fault diagnosis approaches have significantly improved the detection of open-circuit (OC) faults in voltage source inverters (VSIs). However, the diverse operating conditions and noisy environments in electric vehicles (EVs) cause feature variations, rendering challenges in maintaining the high accuracy and robustness of existing methods. This article introduces a robust self-augmented deep learning (DL) method for detecting OC faults in VSIs working at various operation conditions. The proposed method automatically extracts the most relevant statistical features from the measured signals using an automatic feature extraction module (AFEM) based on an improved particle swarm optimization (PSO) algorithm. Then, these optimized features are utilized through an attention-based multi-input convolution neural network (A-MICNN) embedded with a multihead self-attention module (SAM) to effectively capture the local and global characteristics of different OC faults. The A-MICNN architecture effectively processes these multi-inputs by employing shared weights and feature extraction layers to fuse information from statistical features and time-frequency image data. The proposed method is validated using an in-house three-level (3L) F-type inverter setup, achieving a remarkable diagnosis accuracy of 99.74% with high robustness across various speeds, load conditions, and noise levels, providing an efficient solution for EV powertrain fault diagnosis.

• Nguyen, Hoang Nam Du; Huynh, Khang & Robbersmyr, Kjell Gunnar (2025). Robust Multiple-Fault Diagnosis of PMSMs in Dynamic Operations under Imbalanced Datasets. IEEE Transactions on Transportation Electrification. ISSN 2577-4212. 11(5), s. 10714–10726. doi: 10.1109/TTE.2025.3566612. Fulltekst i vitenarkiv Vis sammendrag

  Data-driven models for multiple-fault classifiers of permanent magnet synchronous machines (PMSMs) in electric power trains require historical data at faulty cases in a wide range of operations. However, data in healthy or steady-state operations are much more than the data in the faulty cases, resulting in imbalanced datasets and rendering challenges of expanding the robustness of the data-driven models in dynamic operations. To address these issues, this article proposes a robust two-stage learning scheme for detecting and classifying multiple incipient faults in PMSMs, namely, interturn short circuit (ITSC) local demagnetization (LD) and mixed faults (MFs), under dynamic operations. The first stage focuses on enhancing the effectiveness of anomaly detector in dealing with imbalanced datasets at low-severity faults by integrating prior domain knowledge into anomaly detectors and reducing false alarms using an extreme gradient boosting machine (XGBoost). In the second stage, a training algorithm based on a convolutional neural network (CNN) is introduced to classify multiple faults by studying the appearance of low-severity faults. The performance of the proposed anomaly detector is compared to conventional one-class models, including random forest, local outlier factor, and support vector machine. The accuracy and robustness of the suggested scheme are further evaluated by benchmarking pretrained classifiers such as MobileNetv2, ResNet50, and VGG16, using an imbalanced dataset at different operating conditions from an in-house test setup. A comparative study with a traditional CNN is conducted to explain how the model provides the prediction.

• Nguyen, Thien Phuoc; Huynh, van Khang & Robbersmyr, Kjell Gunnar (2025). Analytical Vibration Model of Stator Inter-Turn Short Circuit and Eccentricity in a Synchronous Generator at Different Loading Conditions, 2025 IEEE Workshop on Electrical Machines Design, Control and Diagnosis - WEMDCD. IEEE (Institute of Electrical and Electronics Engineers). ISSN 9798331520748. doi: 10.1109/WEMDCD61816.2025.11014134. Fulltekst i vitenarkiv Vis sammendrag

  Low-severity stator inter-turn short circuit (ITSC) and orbital eccentricity in synchronous generators (SMs) might not result in clear features or fault characteristics using vibration signals. Further, the effect of load profiles is unavoidable in fault detection of generators. This paper models the effect of the mentioned faults in a SM in various load profiles on the vibration signals on the SM stator surface. The impact of the stator ITSC and eccentricity on vibration harmonics are analyzed using the suggested analytic model based on Fourier expression of magnetic force density. Finally, the results from the suggested analytic model are experimentally validated on an in-house experimental setup.

• Dalal, Anurag; Hagen, Daniel; Robbersmyr, Kjell Gunnar & Knausgård, Kristian Muri (2024). Gaussian Splatting: 3D Reconstruction and Novel View Synthesis: A Review. IEEE Access. 12, s. 96797–96820. doi: 10.1109/ACCESS.2024.3408318. Fulltekst i vitenarkiv
• Eid, Mahmoud Sayed Mahmoud; Ali, Ahmed Mohamed Salem; Huynh, Khang; Robbersmyr, Kjell Gunnar & Pomarnacki, Raimondas (2024). Highly Accurate Diagnosis Scheme of Open-Circuit Faults in a Three-Level Inverter Using Optimized Multi-Classifier Decision Fusion, 2023 24th International Middle East Power System Conference (MEPCON). IEEE (Institute of Electrical and Electronics Engineers). ISSN 9798350358469. doi: 10.1109/MEPCON58725.2023.10462369. Fulltekst i vitenarkiv
• Nguyen, Hoang Nam Du; Huynh, van Khang & Robbersmyr, Kjell Gunnar (2024). Training Scheme for Convolutional Neural Network Based Multiple Fault Classifier of Permanent Magnet Synchronous Motors Under Variable Speed and Load Conditions. Proceedings (International Conference on Electrical Machines) (ICEM). ISSN 2381-4802. doi: 10.1109/ICEM60801.2024.10700238. Fulltekst i vitenarkiv
• Eid, Mahmoud Sayed Mahmoud; Ali, Ahmed Mohamed Salem; Huynh, van Khang & Robbersmyr, Kjell Gunnar (2024). A Few–Shot Open-Circuit Fault Diagnosis of F-Type Inverters Using CGAN-Based Vision Transformer. IEEE Journal of Emerging and Selected Topics in Power Electronics. ISSN 2168-6777. 13. doi: 10.1109/JESTPE.2024.3478378. Fulltekst i vitenarkiv Vis sammendrag

  Multilevel inverters (MLIs) are widely adopted in various industries due to their distinctive features. However, they are susceptible to higher failure rates due to the increased number of components. Deep learning (DL) models are widely used for accurately diagnosing faults in inverters because they effectively extract features automatically. These models work on the hypothesis of the availability of a sufficient number of samples to train the diagnostic models. However, obtaining enough sample data in engineering practice is difficult, especially in fault cases. Therefore, this article proposes a fault diagnosis scheme combining a conditional generative adversarial network (CGAN) and a vision transformer (ViT) for diagnosing opencircuit (OC) faults with few fault samples (few shots). First, the measured signals are converted to time–frequency images. Afterward, CGAN generates new 2-D sample images with data distributions similar to real samples. Finally, the improved ViT uses the original and generated samples to learn and extract local and global features with a multihead self-attention mechanism and classify the faults. The proposed scheme is validated using an experimental setup of the F-type inverter, and the results show that the suggested scheme outperforms the other conventional DL methods with an accuracy of 98.46% using only six samples per class.

• Nguyen, Hoang Nam Du; Huynh, Khang & Robbersmyr, Kjell Gunnar (2024). Classification of Multiple Incipient Faults in Permanent Magnet Synchronous Motors under Variable Speed Conditions Using Transfer Learning. I NN, NN (Red.), 2024 Tenth International Conference on Communications and Electronics (ICCE). IEEE (Institute of Electrical and Electronics Engineers). ISSN 9798350379792. s. 637–642. doi: 10.1109/ICCE62051.2024.10634627. Fulltekst i vitenarkiv
• Pretto, Riccardo; Dayal, Aveen; Haile, Yonas Oghbasilasie; Andrade, José Amendola Netto; Cenkeramaddi, Linga Reddy & Robbersmyr, Kjell Gunnar [Vis alle 7 forfattere av denne artikkelen] (2023). Motor Bearing Fault Classification using Laser Sensor and Light Weight CNN, 2023 IEEE International Symposium on Smart Electronic Systems (iSES). IEEE (Institute of Electrical and Electronics Engineers). ISSN 9798350383249. s. 27–31. doi: 10.1109/iSES58672.2023.00017. Fulltekst i vitenarkiv
• Nguyen, Nam Du Hoang; Huynh, Khang & Robbersmyr, Kjell Gunnar (2023). Current and Stray Flux Sensors for Anomaly Detection in PMSM Drive based on Gradient Boosting Machine. Proceedings of IEEE Sensors. ISSN 1930-0395. doi: 10.1109/SENSORS56945.2023.10325076. Fulltekst i vitenarkiv
• Eid, Mahmoud Sayed Mahmoud; Huynh, Khang; Senanayaka, Jagath Sri Lal; Robbersmyr, Kjell Gunnar & Panadero, Ruben (2023). A two-stage Fault Diagnosis Scheme of Partial Demagnetization Faults in PMSMs with Low Labelled Rates, 2023 IEEE International Electric Machines & Drives Conference (IEMDC). IEEE (Institute of Electrical and Electronics Engineers). ISSN 9798350398991. doi: 10.1109/IEMDC55163.2023.10238950. Fulltekst i vitenarkiv
• Nguyen, Hoang Nam Du; Huynh, van Khang & Robbersmyr, Kjell Gunnar (2023). Early detection of rolling bearing faults using long short-term memory, 2023 6th International Conference on Control, Robotics and Informatics. IEEE (Institute of Electrical and Electronics Engineers). ISSN 9798350323283. s. 46–51. doi: 10.1109/ICCRI58865.2023.00016. Fulltekst i vitenarkiv
• Skår, Eirik Magnus; Kloumann, Jon-Erick; Robbersmyr, Kjell Gunnar & Løvåsen, Torfinn (2023). Anomaly Detection in Robotic Welds - Investigation of LSTM Autoencoder Model Performance. I Zhou, Jing (Red.), 2023 11th International Conference on Control, Mechatronics and Automation (ICCMA). IEEE (Institute of Electrical and Electronics Engineers). ISSN 9798350315684. s. 265–270. doi: 10.1109/ICCMA59762.2023.10374923. Fulltekst i vitenarkiv
• Noorsumar, Gulshan; Rogovchenko, Svitlana; Robbersmyr, Kjell Gunnar; Vysochinskiy, Dmitry & Klausen, Andreas (2023). Development and Extended Validation of a Lumped Parameter Prediction Model for Analysing Injury Parameters in a Vehicle Crash. I Wagner, Gerd; Werner, Frank; Tuncer, Oren & Rango, Floriano De (Red.), Simulation and Modeling Methodologies, Technologies and Applications. SIMULTECH 2021. Springer. ISSN 9783031231490. doi: 10.1007/978-3-031-23149-0_1. Fulltekst i vitenarkiv
• Salata, Gianluca; Cenkeramaddi, Linga Reddy; Huynh, Khang; Robbersmyr, Kjell Gunnar & Jha, Ajit (2023). Time-Frequency Analysis and Fault Prediction of Motor Bearings Using Millimeter-Wave Radar. IEEE Sensors Journal. ISSN 1530-437X. 23(16), s. 18718–18728. doi: 10.1109/JSEN.2023.3292899. Fulltekst i vitenarkiv Vis sammendrag

  Industrial motors are ubiquitous and used extensively in various industrial applications. The aim of increasing productivity, efficiency, lowering down, and mean time between failures while maintaining safety necessitates the continuous condition monitoring of abnormalities, faults in motors, and even predict them as early as possible to maintain their smooth functioning. In this article, we propose and experimentally demonstrate the use of millimeter-wave (mmW) radar for noncontact, nondestructive detection, characterization, and finally prediction of the probability of failure by interrogating its vibrating shaft at a given set of bearing conditions. Precisely, we obtain experimental raw 1-D time series data from mmW radar and transform them into time–frequency representations using wavelet transformation. Then, the features are extracted from the time–frequency representations, specifically the instantaneous velocity trajectory that distinguishes the vibrating shaft due to faulty bearing from healthy bearing. Furthermore, using maximum likelihood estimation (MLE), the velocity profile of the vibrating shaft follows Gaussian distribution from collected data at different motor shaft rotations per second (rps) (5, 15, and 25 Hz) with healthy and faulty bearings. While the mean of distribution for all the cases remains constant, the variance of latter increases with the increase in rps. Finally, the probabilistic approach and the model parameters are exploited to quantify the probability of motor failure in terms of complementary cumulative distribution function (CCDF). It is observed that the CCDF for the faulty case is greater than that of healthy case for all velocities and rps.

• Eid, Mahmoud Sayed Mahmoud; Huynh, Khang; Senanayaka, Jagath Sri Lal & Robbersmyr, Kjell Gunnar (2023). Robust Multiple-Fault Diagnosis of PMSM Drives Under Variant Operations and Noisy Conditions. IEEE Open Journal of the Industrial Electronics Society (OJ-IES). 4, s. 762–772. doi: 10.1109/OJIES.2024.3350443. Fulltekst i vitenarkiv Vis sammendrag

  With the rapid development of industrial applications using permanent magnet synchronous motors (PMSMs) and the Internet of Things, the demand for using robust fault diagnosis methods working in noisy conditions has increased significantly. The current data-driven methods depend mainly on deep learning (DL) models due to the effectiveness of automated feature extraction. However, these models have shallow depths compared with benchmark convolution neural networks, limiting their accuracy in final predictions, and they are established based on the hypothesis that the measured data are noiseless. Despite this, electric machinery is subjected to various noise sources that interfere with measurements during operation. This article proposes a new scheme combining a transfer-learned pretrained residual neural network (ResNet) and supervised machine learning (S-ML) to enhance the performance of DL models in noisy industrial environments. The effectiveness of the proposed scheme is validated using an in-house setup of a PMSM drive with demagnetization and intern-short circuit faults at variant operating conditions. The results show that the proposed method significantly reduced the computational burden by tenfold on average while improving the average accuracy to 96.84% across all the datasets compared with other DL and S-ML methods, with high robustness in noisy working conditions.

• Noorsumar, Gulshan; Rogovchenko, Svitlana; Vysochinskiy, Dmitry & Robbersmyr, Kjell Gunnar (2023). A Numerical Comparison of Lagrange and Kane’s Method for Modeling and Crashworthiness Assessment of a Modified Vehicle. I Wagner, Gerd; Werner, Frank & Rango, Floriano De (Red.), Simulation and Modeling Methodologies, Technologies and Applications. SIMULTECH 2022. Springer. ISSN 9783031438233. s. 156–176. doi: 10.1007/978-3-031-43824-0_9. Fulltekst i vitenarkiv
• Klausen, Andreas; Huynh, Khang & Robbersmyr, Kjell Gunnar (2022). RMS Based Health Indicators for Remaining Useful Lifetime Estimation of Bearings. MIC Journal: Modeling, Identification and Control. ISSN 0332-7353. 43(1), s. 21–38. doi: 10.4173/mic.2022.1.3. Fulltekst i vitenarkiv Vis sammendrag

  Estimating the remaining useful life (RUL) of bearings from healthy to faulty is important for predictive maintenance. The bearing fault severity can be estimated based on the energy or root mean square (RMS) of vibration signals, and a stopping criterion can be set based on a threshold given by an ISO standard. However, the vibration RMS is often not monotonically increasing with damage, which renders a challenge for predicting the RUL. This study proposes a novel method for splitting the vibration signal into multiple frequency bands before RMS calculations to generate multiple health indicators. Monotonic health indicators are identified using the Spearman coefficient, and the RUL is afterward estimated for each indicator using a suitable model and parameter update scheme. Historical failure data is not required to set any parameters. The proposed method is tested with the Paris' law, where parameters are updated by particle filters. Experimental results from two test rigs validate the performance of the proposed method.

• Attestog, Sveinung; Senanayaka, Jagath Sri Lal; Huynh, Van Khang & Robbersmyr, Kjell Gunnar (2022). Robust Active Learning Multiple Fault Diagnosis of PMSM Drives with Sensorless Control under Dynamic Operations and Imbalanced Datasets. IEEE Transactions on Industrial Informatics. ISSN 1551-3203. 19(9), s. 9291–9301. doi: 10.1109/TII.2022.3227628. Fulltekst i vitenarkiv Vis sammendrag

  This paper proposes an active learning scheme to detect multiple faults in permanent magnet synchronous motors in dynamic operations without using historical labelled faulty training data. The proposed method combines the self-supervised anomaly detector based on a local outlier factor (LOF) and a deep Q-network (DQN) supervised reinforcement learner to classify interturn short-circuit, local demagnetisation and mixed faults. The first fault, which is detected by LOF and verified by an expert during maintenance, is used as training data for the DQN classifier. From that point onward, the LOF anomaly detector and DQN fault classifiers are working in tandem in the identification of new faults, which require expert intervention when either of them identifies a fault. The robustness of the scheme against dynamic operations, mixed fault and imbalanced training datasets is validated via a comparative study using stray flux data from an inhouse test setup.

• Attestog, Sveinung; Senanayaka, Jagath Sri Lal; Huynh, Khang & Robbersmyr, Kjell Gunnar (2022). Mixed Fault Classification of Sensorless PMSM Drive in Dynamic Operations Based on External Stray Flux Sensors. Sensors. 22(3). doi: 10.3390/s22031216. Fulltekst i vitenarkiv Vis sammendrag

  This paper aims to classify local demagnetisation and inter-turn short-circuit (ITSC) on position sensorless permanent magnet synchronous motors (PMSM) in transient states based on external stray flux and learning classifier. Within the framework, four supervised machine learning tools were tested: ensemble decision tree (EDT), k-nearest neighbours (KNN), support vector machine (SVM), and feedforward neural network (FNN). All algorithms are trained on datasets from one operational profile but tested on other different operation profiles. Their input features or spectrograms are computed from resampled time-series data based on the estimated position of the rotor from one stray flux sensor through an optimisation problem. This eliminates the need for the position sensors, allowing for the fault classification of sensorless PMSM drives using only two external stray flux sensors alone. Both SVM and FNN algorithms could identify a single fault of the magnet defect with an accuracy higher than 95% in transient states. For mixed faults, the FNN-based algorithm could identify ITSC in parallel-strands stator winding and local partial demagnetisation with an accuracy of 87.1%.

• Noorsumar, Gulshan; Rogovchenko, Svitlana; Robbersmyr, Kjell Gunnar & Vysochinskiy, Dmitry (2022). Vehicle crashworthiness performance in frontal impact: Mathematical model using elastic pendulum. Mechanics research communications. ISSN 0093-6413. 124. doi: 10.1016/j.mechrescom.2022.103954. Fulltekst i vitenarkiv Vis sammendrag

  Vehicle occupant injuries due to collisions cause many fatalities every year. Safe vehicle design plays a critical role in averting serious injuries to occupants and vulnerable road users in the event of a crash. In this paper we study a full frontal vehicle crash against a rigid barrier introducing a Lumped Parameter Model (LPM) inspired by the elastic pendulum motion. The model uses polar coordinates to simplify the problem and the governing equations have been defined using Lagrangian formulation. The Simulink model has been validated against Finite Element (FE) data demonstrating good correlation with pitching angle and maximum crush of the vehicle. These parameters are crucial for designing vehicles which efficiently protect occupants.

• Noorsumar, Gulshan; Rogovchenko, Svitlana; Vysochinskiy, Dmitry & Robbersmyr, Kjell Gunnar (2022). Modeling of Modified Vehicle Crashworthiness using a Double Compound Pendulum. I Wagner, Gerd; Werner, Frank & Rango, Floriano De (Red.), Proceedings of the 12th International Conference on Simulation and Modeling Methodologies, Technologies and Applications. SciTePress. ISSN 9789897585784. s. 102–111. doi: 10.5220/0011306100003274. Fulltekst i vitenarkiv Vis sammendrag

  Vehicle crash modeling has been a challenge for researchers for several decades. Occupant injury prevention and prediction is a critical area within vehicle safety design. The modeling of material failure in structural members during a full frontal crash has been presented in this paper. This study presents a Lumped Parameter Model (LPM) with an elastic double compound pendulum replicating the impact kinematics. The model defined using Lagrangian formulation; presents a novel methodology to represent material fracture caused due to heat affected zones or welding in Ultra High Strength Steels (UHSS) in a non-linear crash event. The material fracture leads to rotation of the vehicle; presented in the form of torsional springs in the LPM developed in this study. The Simulink model has been validated with a finite element simulation and shows good correlation to predict parameters crucial to design for occupant protection in a vehicle crash.

• Husebø, Arild Bergesen; Huynh, Khang; Robbersmyr, Kjell Gunnar & Klausen, Andreas (2022). Classification of Mechanical Fault-Excited Events Based on Frequency. I Sanfilippo, Filippo; Granmo, Ole-Christoffer; Yayilgan, Sule Yildirim & Bajwa, Imran Sarwar (Red.), Intelligent Technologies and Applications: 4th International Conference, INTAP 2021, Grimstad, Norway, October 11–13, 2021, Revised Selected Papers. Springer. ISSN 9783031105241. s. 380–392. doi: 10.1007/978-3-031-10525-8_30. Fulltekst i vitenarkiv Vis sammendrag

  We propose a method for classifying periodic events generated at one or multiple frequencies on any one-dimensional space. This is very useful in problems where you need to find the type of event based on observations of location, e.g. in time. For each frequency, all events are mapped into periodic axes, which are represented independently of each other. Using an expectation-maximization algorithm, we can fit distributions to the events and classify them using maximum likelihood. The proposed method is applied to two mechanical faulty cases: a defect rolling-element bearing, and a gearbox with defect teeth. We show very good classification results in cases of multiple event types of similar frequency, multiple event types of different frequencies, and combinations of the two for artificially generated events.

• Noorsumar, Gulshan; Robbersmyr, Kjell Gunnar; Rogovchenko, Svitlana & Vysochinskiy, Dmitry (2022). An Overview of Data Based Predictive Modeling Techniques Used in Analysis of Vehicle Crash Severity. I Sanfilippo, Filippo; Granmo, Ole-Christoffer; Yayilgan, Sule Yildirim & Bajwa, Imran Sarwar (Red.), Intelligent Technologies and Applications: 4th International Conference, INTAP 2021, Grimstad, Norway, October 11–13, 2021, Revised Selected Papers. Springer. ISSN 9783031105241. s. 355–379. doi: 10.1007/978-3-031-10525-8_28. Fulltekst i vitenarkiv Vis sammendrag

  Accident injury prediction is a crucial constituent to reducing fatalities linked to vehicle crashes. The vehicle development process and road safety planning includes also the injury prediction for occupants and Vulnerable Road Users (VRUs) in a vehicle crash and the identification of the factors responsible for increased traffic collision injuries. This paper reviews the different data-based prediction techniques to modeling a vehicle crash event, crash frequency and crash severity. Machine learning (ML) is a research field which has gained impetus in the recent years and is widely used in different engineering applications; including injury prediction in vehicle collisions. The paper is divided into two major sections; the first section presents an overview of the existing predictive models for estimating injury severity in a crash event to occupants and VRUs and the second section describes the applications of data-based modeling techniques to predict crash frequency in different traffic scenarios. We also discuss possible future applications of data-based modeling techniques in this domain.

• Nyberg, Svein Olav Glesaaen; Robbersmyr, Kjell Gunnar; Holm, Jan Andreas & Sanfilippo, Filippo (2022). Tensile Experiments on Adhesion Between Aluminium Profiles and Glass. I Sanfilippo, Filippo; Granmo, Ole-Christoffer; Yayilgan, Sule Yildirim & Bajwa, Imran Sarwar (Red.), Intelligent Technologies and Applications: 4th International Conference, INTAP 2021, Grimstad, Norway, October 11–13, 2021, Revised Selected Papers. Springer. ISSN 9783031105241. s. 407–418. doi: 10.1007/978-3-031-10525-8_32. Fulltekst i vitenarkiv Vis sammendrag

  In this work, the effects of the adhesion between aluminium profiles and glass are studied from a static tensile perspective. A series of stretch curves are analysed from their derivatives to find their points of float and maximum load bearing. The variable factors are glass type, and type of connection: i.e., edge adhesive, side (fugue) adhesive, and excessive fugue adhesive, for short named fugue-edge. The stretch data imply four quantities to analyse and compare: displacement and load respectively at float and at max load. The results are first compared factor group against factor group. With this tool, only a few significant conclusions may be found. The second comparison is by means of the more robust statistical tools of linear regression and analysis of variance (ANOVA), with conclusions about which factors are significant, and then about the size of the effect on the four variables under study. This forms the basis for a recommendations for how to obtain the strongest possible glass-frame system.

• Knausgård, Kristian Muri; Skar, Siv; Sanfilippo, Filippo; Buldenko, Albert; Lindheim, Henning & Lunde, Jakob E. [Vis alle 8 forfattere av denne artikkelen] (2022). Branch-Manoeuvring Capable Pipe Cleaning Robot for Aquaponic Systems. I Sanfilippo, Filippo; Granmo, Ole-Christoffer; Yayilgan, Sule Yildirim & Bajwa, Imran Sarwar (Red.), Intelligent Technologies and Applications: 4th International Conference, INTAP 2021, Grimstad, Norway, October 11–13, 2021, Revised Selected Papers. Springer. ISSN 9783031105241. s. 107–118. doi: 10.1007/978-3-031-10525-8_9. Fulltekst i vitenarkiv Vis sammendrag

  Aquaponic systems are engineered ecosystems combining aquaculture and plant production. Nutrient rich water is continuously circulating through the system from aquaculture tanks. A biofilter with nitrifying bacteria breaks down fish metabolism ammonia into nitrite and nitrate, which plants and makes the aquaculture wastewater into valued organic fertiliser for the plants, containing essential macro and micro elements. At the same time, the plants are cleaning the water by absorbing ammonia from the fish tanks before it reaches dangerous levels for the aquatic animals. In principle, the only external input is energy, mainly in the form of light and heat, but fish food is also commonly provided. Growing fish food is potentially feasible in a closed loop system, hence aquaponic systems can possibly be an important source of proteins and other important nutrition when, for example, colonising other planets in the future. Fully autonomous aquaponic systems are currently not available. This work aims at minimising manual labour related to cleaning pipes for water transport. The cleaning process must be friendly to both plants and aquatic animals. Hence, in this work, pure mechanical cleaning is adopted. A novel belt-driven continuum robot capable of travelling through small/medium diameter pipes and manoeuvring branches and bends, is designed and tested. The robot is modular and can be extended with different cleaning modules through an interface providing CAN-bus network and electric power. The flexible continuum modules of the robot are characterised. Experimental results demonstrate that the robot is able to travel through pipes with diameters varying from 50 mm to 75 mm, and also capable of handling T-branches of up to 90∘.

• Sanfilippo, Filippo; Thorstensen, Rein Terje; Jha, Ajit; Jiang, Zhiyu & Robbersmyr, Kjell Gunnar (2022). A Perspective Review on Digital Twins for Roads, Bridges, and Civil Infrastructures. I -, IEEE (Red.), IEEE 2022 3rd International Conference on Electrical, Computer, Communications and Mechatronics Engineering (ICECCME). IEEE (Institute of Electrical and Electronics Engineers). ISSN 9781665470957. doi: 10.1109/ICECCME55909.2022.9988693. Fulltekst i vitenarkiv
• Ali, Ahmed Mohamed Salem; Huynh, Van Khang; Jiya, Immanuel Ninma & Robbersmyr, Kjell Gunnar (2022). Hybrid Three-Phase Transformer-Based Multilevel Inverter with Reduced Component Count. IEEE Access. 10, s. 47754–47763. doi: 10.1109/ACCESS.2022.3171849. Fulltekst i vitenarkiv Vis sammendrag

  The topology of the static synchronous compensator of reactive power for a low-voltage three-phase utility grid capable of asymmetric reactive power compensation in grid phases has been proposed and analysed. It is implemented using separate, independent cascaded H-bridge multilevel inverters for each phase. Every inverter includes two H-bridge cascades. The first cascade operating at grid frequency is implemented using thyristors, and the second one—operating at high frequency is based on the high-speed MOSFET transistors. The investigation shows that the proposed compensator is able to compensate the reactive power in a low-voltage three-phase grid when phases are loaded by highly asymmetrical reactive loads and provides up to three times lower power losses in the compensator as compared with the situation when the compensator is based on the conventional three-level inverters implemented using IGBT transistors.

• Salem, Ahmed; Huynh, Khang & Robbersmyr, Kjell Gunnar (2021). Four-level three-phase inverter with reduced component count for low and medium voltage applications. IEEE Access. 9, s. 35151–35163. doi: 10.1109/ACCESS.2021.3062110. Fulltekst i vitenarkiv
• Husebø, Arild Bergesen; Klausen, Andreas; Huynh, Khang & Robbersmyr, Kjell Gunnar (2021). A Simple Time Domain-Based Method for Estimating the Resonance Frequency of a Bearing, 2021 24th International Conference on Electrical Machines and Systems (ICEMS). IEEE (Institute of Electrical and Electronics Engineers). ISSN 9788986510218. doi: 10.23919/ICEMS52562.2021.9634452. Fulltekst i vitenarkiv
• Noorsumar, Gulshan; Vysochinskiy, Dmitry; Englund, Even; Robbersmyr, Kjell Gunnar & Rogovchenko, Svitlana (2021). Effect of welding and heat treatment on the properties of UHSS used in automotive industry. The European Physical Journal Conferences. ISSN 2101-6275. 250. doi: 10.1051/epjconf/202125005015. Fulltekst i vitenarkiv Vis sammendrag

  This paper deals with the undesired effects of the heat treatments on the mechanical properties of (UHSS) Ultra High Strength Steel used nowadays in automotive industry to improve crashworthiness performance of vehicles. The UHSS specimens were extracted from certain parts of the car body and subjected to different heat treatments. Four types of specimens were tested: untreated, welded with metal inert gas welding, heat treated at 800 °C, and heat treated at 1250 °C. All heat-treated specimens showed dramatically reduced values of strength. The results suggest that it is important to follow the official repair manuals avoiding unnecessary welding and improper heat treatments of UHSS. The experiments provide the data necessary for constructing a constitutive model and performing a finite-element analysis of improperly repaired UHSS parts.

• Noorsumar, Gulshan; Rogovchenko, Svitlana; Robbersmyr, Kjell Gunnar & Vysochinskiy, Dmitry (2021). Mathematical models for assessment of vehicle crashworthiness: a review. International Journal of Crashworthiness. ISSN 1358-8265. 27(5). doi: 10.1080/13588265.2021.1929760. Fulltekst i vitenarkiv Vis sammendrag

  This article reviews approaches to mathematical modeling of a vehicle crash. The growing focus on vehicle and occupant safety in car crashes has triggered the need to study vehicle crashworthiness in the initial stages of vehicle development. The major motivation for this work is to support vehicle crashworthiness design during the product development process.The article is divided into two parts; the first one overviews existing mathematical models used to solve engineering problems. The second part describes modeling strategies applied for replicating non-linear vehicle crash event and occupant kinematics in an occupant protection loadcase. We also highlight alternative modeling strategies using hybrid modeling techniques aimed at the improvement of the vehicle development process.

• Attestog, Sveinung; Huynh, Khang & Robbersmyr, Kjell Gunnar (2021). Field reconstruction for modeling multiple faults in permanent magnet synchronous motors in transient states. IEEE Access. 9, s. 127131–127140. doi: 10.1109/ACCESS.2021.3112224. Fulltekst i vitenarkiv
• Ali, Ahmed Mohamed Salem; Huynh, Van Khang & Robbersmyr, Kjell Gunnar (2021). Novel Transformer-Based Six-Level Inverter with Low Component Count. I Bossche, Alex (Red.), 2021 23rd European Conference on Power Electronics and Applications (EPE'21 ECCE Europe). IEEE (Institute of Electrical and Electronics Engineers). ISSN 9789075815375. Fulltekst i vitenarkiv
• Senanayaka, Jagath Sri Lal; Huynh, Van Khang & Robbersmyr, Kjell Gunnar (2021). CNN based Gearbox Fault Diagnosis and Interpretation of Learning Features, IEEE 30th International Symposium on Industrial Electronics (ISIE). IEEE (Institute of Electrical and Electronics Engineers). ISSN 9781728156354. doi: 10.1109/ISIE45552.2021.9576257. Fulltekst i vitenarkiv
• Noorsumar, Gulshan; Rogovchenko, Svitlana; Robbersmyr, Kjell Gunnar; Vysochinskiy, Dmitry & Klausen, Andreas (2021). A Novel Technique for Modeling Vehicle Crash using Lumped Parameter Models, 11th International Conference on Simulation and Modeling Methodologies, Technologies and Applications (SIMULTECH 2021). SciTePress. ISSN 9789897585289. s. 62–70. doi: 10.5220/0010529200620070. Fulltekst i vitenarkiv Vis sammendrag

  This paper presents a novel technique for modeling a full frontal vehicle crash. The crash event is divided into two phases; the first until maximum crush and the second part when the vehicle starts pitching forward. This novel technique will help develop a three degrees of freedom (DOF) lumped parameter model (LPM) for crash and support in the vehicle development process. The paper also highlights the design process for reducing vehicle pitching in occupant protection load cases. The model has been validated against a finite element (FE) simulation of a full frontal crash of a Chevrolet Silverado developed by the National Highway Traffic Safety Administration (NHTSA), and the LPM shows good correlation with the FE test data.

• Jakobsen, Mette Mo; Tørdal, Sondre Sanden; Klausen, Andreas; Robbersmyr, Kjell Gunnar & Tyapin, Ilya (2021). Product development and project-based learning in Mechatronics, in the context of digitization and sustainability. Nordic Journal of STEM Education. 5(1). doi: 10.5324/njsteme.v5i1.3946. Fulltekst i vitenarkiv
• Wæhle, Andreas; Rogne, Nora Kvingedal; Westbye, Stian; Nilsen, Jørgen; Tollevik, Sigurd & Robbersmyr, Kjell Gunnar (2021). How to succeed with volunteers on the basis of commitment and ownership. Nordic Journal of STEM Education. 5(1). doi: 10.5324/njsteme.v5i1.3906. Fulltekst i vitenarkiv
• Wæhle, Andreas; Nilsen, Jørgen; Tollevik, Sigurd; Rogne, Nora Kvingedal; Westbye, Stian & Robbersmyr, Kjell Gunnar (2021). The culture of product development in student organisations. Nordic Journal of STEM Education. 5(1). doi: 10.5324/njsteme.v5i1.3920. Fulltekst i vitenarkiv
• Salem, Ahmed; Huynh, Khang; Robbersmyr, Kjell Gunnar; Norambuena, Margarita & Rodriguez, Jose (2020). Novel Three-Phase Multilevel Inverter with Reduced Components for Low- and High-Voltage Applications. IEEE transactions on industrial electronics (1982. Print). ISSN 0278-0046. doi: 10.1109/TIE.2020.2998752. Fulltekst i vitenarkiv
• Hemmer, Martin; Klausen, Andreas; Huynh, Khang; Robbersmyr, Kjell Gunnar & Waag, Tor Inge (2020). Health Indicator for Low-speed Axial Bearings using Variational Autoencoders. IEEE Access. 8, s. 35842–35852. doi: 10.1109/ACCESS.2020.2974942. Fulltekst i vitenarkiv Vis sammendrag

  This paper proposes a method for calculating a health indicator (HI) for low-speed axial rolling element bearing (REB) health assessment by utilizing the latent representation obtained by variational inference using Variational Autoencoders (VAEs), trained on each speed reference in the dataset. Further, versatility is added by conditioning on the speed, extending the VAE to a conditional VAE (CVAE), thereby incorporating all speeds in a single model. Within the framework, the coefficients of autoregressive (AR) models are used as features. The dimensionality reduction inherent in the proposed method lowers the need of expert knowledge to design good condition indicators. Moreover, the suggested methodology allows for setting the probability of false alarms when encoding new data points to the latent variable space using the trained model. The effectiveness of the proposed method is validated based on two different datasets: from a workshop test of an offshore drilling machine and from an in-house test rig for axial bearings. In both datasets, the HI is exceeding the warning and alarm levels with a probability of false alarm (PFA) of 10 -6 , and the method is most effective at lower shaft speeds.

• Husebø, Arild Bergesen; Kandukuri, Surya Teja; Klausen, Andreas; Huynh, Khang & Robbersmyr, Kjell Gunnar (2020). Rapid Diagnosis of Induction Motor Electrical Faults using Convolutional Autoencoder Feature Extraction. Proceedings of the European Conference of the Prognostics and Health Management Society (PHME). 5(1). Fulltekst i vitenarkiv
• Senanayaka, Jagath Sri Lal; Huynh, Khang & Robbersmyr, Kjell Gunnar (2020). Towards Self-Supervised Feature Learning for Online Diagnosis of Multiple Faults in Electric Powertrains. IEEE Transactions on Industrial Informatics. ISSN 1551-3203. doi: 10.1109/TII.2020.3014422. Fulltekst i vitenarkiv Vis sammendrag

  This paper proposes a novel online fault diagnosis scheme for industrial powertrains without using historical faulty or labelled training data. The proposed method combines one-class support vector machine (SVM)-based anomaly detection and supervised convolutional neural network (CNN) algorithms to online detect multiple faults and fault severities under variable speeds and loads. The one-class SVM algorithm is to derive a score for defining faults or health classes in the first stage, and the resulting health classes are used as the training data for the CNN-based classifier in the second stage. Within this framework, the self-supervised learning of the proposed CNN algorithm allows the online diagnosis scheme to learn features based on the latest data. The effectiveness of the scheme is validated via a comparison study using experimental data from an in-house test setup. Finally, the online implementation of the proposed scheme on the test setup is briefly introduced.

• Salem, Ahmed; Huynh, Khang; Robbersmyr, Kjell Gunnar; Norambuena, Margarita & Rodríguez, José (2020). Voltage Source Multilevel Inverters with Reduced Device Count: Topological Review and Novel Comparative Factors. IEEE transactions on power electronics. ISSN 0885-8993. 36(3), s. 2720–2747. doi: 10.1109/TPEL.2020.3011908. Fulltekst i vitenarkiv Vis sammendrag

  Multilevel inverters (MLIs) have gained increasing interest for advanced energy-conversion systems due to their features of high-quality produced waveforms, modularity, transformerless operation, voltage, and current scalability, and fault-tolerant operation. However, these merits usually come with the cost of a high number of components. Over the past few years, proposing new MLIs with a lower component count has been one of the most active topics in power electronics. The first aim of this article is to update and summarize the recently developed multilevel topologies with a reduced component count, based on their advantages, disadvantages, construction, and specific applications. Within the framework, both single-phase and three-phase topologies with symmetrical and asymmetrical operations are taken into consideration via a detailed comparison in terms of the used component count and type. The second objective is to propose a comparative method with novel factors to take component ratings into account. The effectiveness of the proposed method is verified by a comparative study.

• Attestog, Sveinung; Huynh, Khang & Robbersmyr, Kjell Gunnar (2020). Improved Quadratic Time-frequency Distributions for Detecting Inter-turn Short Circuits of PMSMs in Transient States. I Liu, Yujing (Red.), Proc. 2020 International Conference on Electrical Machines (ICEM). IEEE (Institute of Electrical and Electronics Engineers). ISSN 9781728199450. s. 1461–1467. doi: 10.1109/ICEM49940.2020.9271050. Fulltekst i vitenarkiv
• Gåsvik, Kjell; Robbersmyr, Kjell G. & Vadseth, Trond Kjell (2020). Online dimensional control of rolled steel profiles using projected fringes. The International Journal of Advanced Manufacturing Technology. ISSN 0268-3768. 107(3-4), s. 1725–1730. doi: 10.1007/s00170-020-05078-3. Fulltekst i vitenarkiv Vis sammendrag

  Fringe projection is a versatile method for mapping the topography of surfaces. In this paper, it is used to measure the defects on the head of railroad rails while the rails are moving. Railroad rails are made by hot rolling. The quality of the finished product is generally good, but surface texture will deteriorate with increasing temperature. A method for online inspection therefore is very desirable. In the present experiment, dimensional inspection of the railroad rails was made online while moving at a speed of 1–2 m/s. Therefore, it is important to minimize the registration time. To achieve this, we apply a method of fringe location with sub-pixel accuracy that requires only a single exposure for each registration.

• Gåsvik, Kjell; Robbersmyr, Kjell Gunnar & Vadseth, Trond Kjell (2020). Phase error analysis of clipped waveforms in surface topography measurement using projected fringes. Optics Communications. ISSN 0030-4018. 483(126601). doi: 10.1016/j.optcom.2020.126601. Fulltekst i vitenarkiv Vis sammendrag

  When working with the method of projected fringes outside the optical laboratory one often encounters the problem of uncontrollable ambient light. This might cause saturation of the camera which in turn results in clipping of the fringes. Since standard theories describing phase-shifting techniques assume the projected fringes to be purely sinusoidal, such clipping will result in measurement error. In this paper a detailed analysis of this problem is given, and relations between phase errors, the amount of fringe clipping and the number of phase steps are found. Moreover, the phase difference between the clipped and the unclipped fringes is described. This investigation is based on Fourier- and phasor analysis.

• Noorsumar, Gulshan; Robbersmyr, Kjell Gunnar; Rogovchenko, Svitlana & Vysochinskiy, Dmitry (2020). Crash Response of a Repaired Vehicle-Influence of Welding UHSS Members. SAE technical paper series. ISSN 0148-7191. 1- 2020(April), s. 1–5. doi: 10.4271/2020-01-0197. Fulltekst i vitenarkiv Vis sammendrag

  Automakers generally recommend not to weld structural parts after a vehicle crash, and these should be replaced as a whole part in case of a crash event. Sectioning of these members is also not recommended and use of the repair manual is mandatory in case of fracture of such parts. However, repair shops may not adhere to these instructions and use incorrect repair procedures on these members which would modify their strength properties. This study analyses the impact of welding structural members in a vehicle like the A-pillar which use Ultra-High Strength Steels (UHSS) for reducing the weight of the vehicle and improving the crashworthiness of the structure. The research conducted in this paper highlights the differences in the crash performance of a repaired vehicle as opposed to baseline injury values for the vehicle. The performance of the modified vehicle when tested for different loadcases shows reduced crash performance as compared to the baseline performance and it can be concluded that welding or sectioning the UHSS parts would influence the crashworthiness of a vehicle. This paper only focuses on structural integrity of the repaired vehicle in a crash event. The performance of the vehicle in occupant injury is kept out of scope for this study.

• Klausen, Andreas; Kalaoja, Johannes; Kandukuri, Surya Teja & Robbersmyr, Kjell Gunnar (2020). Sensitivity Analysis Of Online Oil Quality Monitoring For Early Detection Of Water Ingress In Marine Propulsion Systems. Proceedings of the European Conference of the Prognostics and Health Management Society (PHME). 5(1). doi: 10.36001/phme.2020.v5i1.1268. Fulltekst i vitenarkiv Vis sammendrag

  Gearboxes are critical equipment in the vessel propulsion system. Lubrication contamination caused by water ingress or condensation is one of the major failure modes in marine gearboxes leading to accelerated aging of lubricant, resulting in accelerated wear of gearbox components such as bearing and gears. This article presents a systematic evaluation of moisture ingress sensitivity of three commercial online oil quality monitoring sensors. A laboratory setup consisting of an industrial multi-stage planetary gearbox is utilised for the experimental studies. Furthermore, the gearbox vibration signatures are analysed to investigate the sensitivity of vibration measurements to water ingress to determine the need for oil quality monitoring when considering this specific failure mode.

• Hemmer, Martin; Robbersmyr, Kjell Gunnar; Waag, Tor Inge; Albrigtsen, Rolf; Pedersen, Torfinn & Meyer, Thomas [Vis alle 7 forfattere av denne artikkelen] (2020). Rib-roller wear in tapered rolling element bearings: Analysis and development of test rig for condition monitoring. Lecture Notes in Mechanical Engineering. ISSN 2195-4356. s. 750–757. doi: 10.1007/978-3-030-48021-9_83. Fulltekst i vitenarkiv Vis sammendrag

  Rolling Element Bearings (REBs) are present in virtually all machines with moving or rotating parts, and are vital for proper performance and safe operation. Condition Monitoring (CM) of bearings often receive particular interest, as this component group rarely reach design lifetime and hence is responsible for unplanned machine downtime. Unplanned maintenance can represent a large cost which motivates development of improved CM methods for implementation of advanced maintenance regimes. Based on observations of a used bearing from an offshore drilling machine, wear on roller ends in the rib-roller contact area was identified as an area of interest for future research. A test rig for creating and observing accelerated roller end damage is developed, intended for use with vibration and Acoustic Emission (AE) sensors. In addition to normal continuous rotation of the bearing, the test rig is also designed with performing oscillation motion tests in mind. This mode of operation is of interest to manufacturers and end users of cranes and winches with heave compensation. Plans and challenges for future work are also discussed, in conjunction with the experimental setup.

• Klausen, Andreas; Huynh, Khang & Robbersmyr, Kjell Gunnar (2019). Multi-band identification for enhancing bearing fault detection in variable speed conditions. Mechanical systems and signal processing. ISSN 0888-3270. 139. doi: 10.1016/j.ymssp.2019.106422. Fulltekst i vitenarkiv
• Hemmer, Martin; Klausen, Andreas; Huynh, Khang; Robbersmyr, Kjell Gunnar & Waag, Tor Inge (2019). Simulation-driven Deep Classification of Bearing Faults from Raw Vibration Data. International Journal of Prognostics and Health Management. 10(Special Issue on Deep Learning and Emerging Analytics). Fulltekst i vitenarkiv Vis sammendrag

  The industry is moving towards maintenance strategies that consider component health, which require extensive collection and analysis of data. Condition monitoring methods that require manual feature extraction and analysis, become infeasible on an industrial scale. Machine learning algorithms can be used to automatically detect and classify faults, however, obtaining sufficient data for training is required for deep learning and other data-driven classification approaches. Data from healthy machine operation is generally available in abundance, while data from representative fault- and operating conditions is limited. This limits both development and deployment of deep learning-based CM systems on an industrial scale. This paper addresses both the challenges of automated analysis and lack of training data. A deep learning classifier architecture utilizing 1-dimensional dilated convolutions is proposed. Dilation of the convolution kernel allows for analysis of raw vibration signals while simultaneously maintaining the receptive field of the classifier enough to capture temporal patterns. The proposed method performs classification in time domain on signal segments of 1 second or shorter. With knowledge of the bearing specification, artificial vibration signals with similar characteristics as an actual bearing fault can be created. In this work, generated fault signals are combined with healthy operational data to obtain training data for a deep classifier. Parameters of the vibration model is chosen as distributions rather than fixed values. By using a range parameters in the vibration model, the classifier learns to recognize temporal features from the training data that generalize to unseen data. The effectiveness of the proposed method is demonstrated by training classifiers on generated data and testing on real signals from faulty bearings at both low and high speed. One dataset containing seeded faults and three run-to-failure tests are used for the demonstration.

• Attestog, Sveinung; Huynh, Khang & Robbersmyr, Kjell Gunnar (2019). Detecting Eccentricity and Demagnetization Fault of Permanent Magnet Synchronous Generators in Transient State, 22nd International Conference on Electrical Machines and Systems (ICEMS) 2019. IEEE (Institute of Electrical and Electronics Engineers). ISSN 9781728133980. doi: 10.1109/ICEMS.2019.8921753. Fulltekst i vitenarkiv
• Klausen, Andreas & Robbersmyr, Kjell Gunnar (2019). Cross-correlation of whitened vibration signals for low-speed bearing diagnostics. Mechanical systems and signal processing. ISSN 0888-3270. 118(1), s. 226–244. doi: 10.1016/j.ymssp.2018.08.048. Fulltekst i vitenarkiv
• Ali, Ahmed Mohamed Salem; Huynh, Khang & Robbersmyr, Kjell Gunnar (2019). New Multilevel Inverter Topology with Reduced Component Count, EPE'19 ECCE Europe. IEEE (Institute of Electrical and Electronics Engineers). ISSN 9789075815313. doi: 10.23919/EPE.2019.8915400. Fulltekst i vitenarkiv
• Ali, Ahmed Mohamed Salem; Huynh, Khang; Robbersmyr, Kjell Gunnar; Norambuena, Margarita & Rodriguez, Jose (2019). Novel Three-Phase Multi-Level Inverter with Reduced Components, Proceeding 45th Annual Conference of the IEEE Industrial Electronics Society - IECON 2019. IEEE (Institute of Electrical and Electronics Engineers). ISSN 9781728148786. s. 6501–6506. doi: 10.1109/IECON.2019.8927732. Fulltekst i vitenarkiv
• Kandukuri, Surya Teja; Huynh, Khang & Robbersmyr, Kjell Gunnar (2019). Diagnostics of stator winding failures in wind turbine pitch motors using Vold-Kalman filter, Proceeding 45th Annual Conference of the IEEE Industrial Electronics Society - IECON 2019. IEEE (Institute of Electrical and Electronics Engineers). ISSN 9781728148786. s. 5992–5997. doi: 10.1109/IECON.2019.8926983. Fulltekst i vitenarkiv Vis sammendrag

  Pitch systems are among the most failure-prone components in wind turbines. Winding failures in pitch motors are common due to high start-up loads and poor ventilation. This article presents a diagnostics scheme that can detect the stator winding failures in the pitch motors under time-varying speed and load conditions. The proposed approach based on three-phase motor currents can be directly integrated into the motor drive and can be used for induction as well as permanent magnet synchronous machines. The extended Park's vector calculated on the motor currents is order tracked based on the supply frequency from the drive using Vold-Kalman filter. The approach is shown to be robust under arbitrary load and speed variations in a laboratory setup of a pitch drive.

• Skylvik, Audun; Robbersmyr, Kjell Gunnar & Huynh, Khang (2019). Data-driven Fault Diagnosis of Induction Motors Using a Stacked Autoencoder Network, 22nd International Conference on Electrical Machines and Systems (ICEMS) 2019. IEEE (Institute of Electrical and Electronics Engineers). ISSN 9781728133980. doi: 10.1109/ICEMS.2019.8921738. Fulltekst i vitenarkiv Vis sammendrag

  Current signatures from an induction motor are normally used to detect anomalies in the condition of the motor based on signal processing techniques. However, false alarms might occur if using signal processing analysis alone since missing frequencies associated with faults in spectral analyses does not guarantee that a motor is fully healthy. To enhance fault diagnosis performance, this paper proposes a machine learning based method using in-built motor currents to detect common faults in induction motors, namely inter-turn stator winding-, bearing- and broken rotor bar faults. This approach utilizes single-phase current data, being pre-processed using Welch’s method for spectral density estimation. Further, several deep learning features are extracted using stacked autoencoder networks. The proposed scheme can predict the faults at high accuracy while reducing expertise demand. The proposed method is validated on an in-house laboratory setup under variable speed and load conditions, showing a classification accuracy over 95%. Furthermore, multiple support vector machine and k-nearest neighbor classifiers are tested with the same data for a comparative study. Finally, the performance of different fault classifiers are evaluated based on accuracy, computational burden and classification consistency.

• Kandukuri, Surya Teja; Senanayaka, Jagath Sri Lal; Huynh, Khang & Robbersmyr, Kjell Gunnar (2019). A Two-Stage Fault Detection and Classification Scheme for Electrical Pitch Drives in Offshore Wind Farms Using Support Vector Machine. IEEE transactions on industry applications. ISSN 0093-9994. 55(5), s. 5109–5118. doi: 10.1109/TIA.2019.2924866. Fulltekst i vitenarkiv Vis sammendrag

  Pitch systems are one of the components with the most frequent failure in wind turbines. This article presents a two-stage fault detection and classification scheme, for electric motor drives in wind turbine pitch systems. The presented approach is suitable for application in offshore wind farms, with electric pitch systems driven by induction motors as well as permanent magnet synchronous motors. The adopted strategy utilizes three-phase motor current sensing at the pitch drives for fault detection and only when a fault condition is detected at this stage, features extracted from the current signals are transmitted to a support vector machine classifier located centrally to the wind farm. The proposed method is validated in a laboratory setup of the pitch drive.

• Kandukuri, Surya Teja; Huynh, Khang & Robbersmyr, Kjell Gunnar (2019). Diagnosis of inverter-fed induction motors in short time windows using physics-assisted deep learning framework, 2019 IEEE 12th International Symposium on Diagnostics for Electrical Machines, Power Electronics and Drives (SDEMPED). IEEE (Institute of Electrical and Electronics Engineers). ISSN 9781728118321. s. 264–270. doi: 10.1109/DEMPED.2019.8864889. Fulltekst i vitenarkiv
• Attestog, Sveinung; Huynh, Khang & Robbersmyr, Kjell Gunnar (2019). Modelling Demagnetized Permanent Magnet Synchronous Generators using Permeance Network Model with Variable Flux Sources, 2019 22nd International Conference on the Computation of Electromagnetic Fields (COMPUMAG). IEEE (Institute of Electrical and Electronics Engineers). ISSN 9781728155920. doi: 10.1109/COMPUMAG45669.2019.9032791. Fulltekst i vitenarkiv
• Senanayaka, Jagath Sri Lal; Huynh, Khang & Robbersmyr, Kjell Gunnar (2018). Multiple Classifiers and Data Fusion for Robust Diagnosis of Gearbox Mixed Faults. IEEE Transactions on Industrial Informatics. ISSN 1551-3203. 15(8), s. 4569–4579. doi: 10.1109/TII.2018.2883357. Fulltekst i vitenarkiv Vis sammendrag

  Detection and isolation of single and mixed-faults in a gearbox are very important to enhance the system reliability, lifetime, and service availability. This paper proposes a hybrid learning algorithm, consisting of Multilayer Perceptron (MLP)- and Convolutional Neural Network (CNN)-based classifiers, for diagnosis of gearbox mixed faults. Domain knowledge features are required to train the MLP classifier, while the CNN classifier can learn features itself, allowing to reduce the required knowledge features for the counterpart. Vibration data from an experimental setup with gearbox mixed faults is used to validate the effectiveness of the algorithms and compare them with conventional methods. The comparative study shows that accuracies and robustness of the individual MLP- and CNN-algorithms are better than those of the compared methods and can be significantly improved using data fusion at the feature level. Furthermore, the robustness of the algorithm is secured under noises by combining the results of individual classifiers.

• Hemmer, Martin; Huynh, Khang; Robbersmyr, Kjell Gunnar; Waag, Tor Inge & Meyer, Thomas (2018). Fault Classification of Axial and Radial Roller Bearings Using Transfer Learning through a Pretrained Convolutional Neural Network. Designs. doi: 10.3390/designs2040056. Fulltekst i vitenarkiv Vis sammendrag

  Detecting bearing faults is very important in preventing non-scheduled shutdowns, catastrophic failures, and production losses. Localized faults on bearings are normally detected based on characteristic frequencies associated with faults in time and frequency spectra. However, missing such characteristic frequency harmonics in a spectrum does not guarantee that a bearing is healthy, or noise might produce harmonics at characteristic frequencies in the healthy case. Further, some defects on roller bearings could not produce characteristic frequencies. To avoid misclassification, bearing defects can be detected via machine learning algorithms, namely convolutional neural network (CNN), support vector machine (SVM), and sparse autoencoder-based SVM (SAE-SVM). Within this framework, three fault classifiers based on CNN, SVM, and SAE-SVM utilizing transfer learning are proposed. Transfer of knowledge is achieved by extracting features from a CNN pretrained on data from the imageNet database to classify faults in roller bearings. The effectiveness of the proposed method is investigated based on vibration and acoustic emission signal datasets from roller bearings with artificial damage. Finally, the accuracy and robustness of the fault classifiers are evaluated at different amounts of noise and training data.

• Senanayaka, Jagath Sri Lal; Huynh, Khang & Robbersmyr, Kjell Gunnar (2018). Autoencoders and Data Fusion Based Hybrid Health Indicator for Detecting Bearing and Stator Winding Faults in Electric Motors. I Kim, Chang-Eob (Red.), 2018 21st International Conference on Electrical Machines and Systems (ICEMS). IEEE (Institute of Electrical and Electronics Engineers). ISSN 9788986510201. s. 531–536. doi: 10.23919/ICEMS.2018.8549064. Fulltekst i vitenarkiv
• Senanayaka, Jagath Sri Lal; Huynh, Khang & Robbersmyr, Kjell Gunnar (2018). Autoencoders and Recurrent Neural Networks Based Algorithm for Prognosis of Bearing Life. I Kim, Chang-Eob (Red.), 2018 21st International Conference on Electrical Machines and Systems (ICEMS). IEEE (Institute of Electrical and Electronics Engineers). ISSN 9788986510201. s. 537–542. doi: 10.23919/ICEMS.2018.8549006. Fulltekst i vitenarkiv
• Kandukuri, Surya Teja; Robbersmyr, Kjell Gunnar & Karimi, Hamid Reza (2018). Toward farm-level health management of wind turbine systems: status and scope for improvements. I Karimi, Hamid Reza (Red.), Structural Control and Fault Detection of Wind Turbine Systems. IET Digital Library. ISSN 9781785613944. s. 149–168. doi: 10.1049/PBPO117E_ch6. Fulltekst i vitenarkiv
• Kandukuri, Surya Teja; Huynh, Khang & Robbersmyr, Kjell Gunnar (2018). Multi-Component Fault Detection in Wind Turbine Pitch Systems Using Extended Park's Vector and Deep Autoencoder Feature Learning. I Kim, Chang-Eob (Red.), 2018 21st International Conference on Electrical Machines and Systems (ICEMS). IEEE (Institute of Electrical and Electronics Engineers). ISSN 9788986510201. s. 1002–1007. doi: 10.23919/ICEMS.2018.8549293. Fulltekst i vitenarkiv

Se alle arbeider i NVA

• Davidsen, Jan Grimsrud; Hellum, Vidar; Robbersmyr, Kjell Gunnar; Jakobsen, Mette Mo; Sjursen, Aslak Darre & Frøiland, Eirik [Vis alle 7 forfattere av denne artikkelen] (2026). Status for utvikling av automatisk sortering av fisk ved hjelp av kunstig intelligens. doi: https:/www.fhf.no/nyheter-og-arrangementer/arrangementer/pukkellaks-2026/. Fulltekst i vitenarkiv
• Davidsen, Jan Grimsrud; Hellum, Vidar; Robbersmyr, Kjell Gunnar; Jakobsen, Mette Mo; Sjursen, Aslak Darre & Frøiland, Eirik [Vis alle 7 forfattere av denne artikkelen] (2025). Status of development ofautomatic sorting of fish using AI. Fulltekst i vitenarkiv
• Eid, Mahmoud Sayed Mahmoud; Huynh, van Khang; Senanayaka, Jagath Sri Lal; Robbersmyr, Kjell Gunnar & Panadero, Ruben (2023). A two-stage Fault Diagnosis Scheme of Partial Demagnetization Faults in PMSMs with Low Labelled Rates. Fulltekst i vitenarkiv
• Nguyen, Hoang Nam Du; Huynh, van Khang & Robbersmyr, Kjell Gunnar (2023). Current and stray flux sensors for anomaly detection in PMSM drive based on gradient boosting machine. Fulltekst i vitenarkiv
• Pretto, Riccardo; Dayal, Aveen; Haile, Yonas Oghbasilasie; Andrade, José Amendola Netto; Cenkeramaddi, Linga Reddy & Robbersmyr, Kjell Gunnar [Vis alle 7 forfattere av denne artikkelen] (2023). Motor Bearing Fault Classification using Laser Sensor and Light Weight CNN. Fulltekst i vitenarkiv
• Nyberg, Svein Olav Glesaaen; Robbersmyr, Kjell Gunnar; Holm, Jan Andreas & Sanfilippo, Filippo (2021). Effects of Material Selection on Adhesion between glass and aluminium pre-recorded video for digital attendees. Fulltekst i vitenarkiv
• Sanfilippo, Filippo; Thorstensen, Rein Terje; Jha, Ajit; Jiang, Zhiyu & Robbersmyr, Kjell Gunnar (2021). A Perspective of Road Infrastructure and Bridges Digital Twins. Fulltekst i vitenarkiv
• Husebø, Arild Bergesen; Huynh, Khang; Robbersmyr, Kjell Gunnar & Klausen, Andreas (2021). Classification of Mechanical Fault-Excited Events based on Frequency. Fulltekst i vitenarkiv
• Noorsumar, Gulshan; Rogovchenko, Svitlana; Robbersmyr, Kjell Gunnar; Vysochinskiy, Dmitry & Klausen, Andreas (2021). A Novel Technique for Modeling Vehicle Crash using Lumped Parameter Models. Fulltekst i vitenarkiv
• Noorsumar, Gulshan; Robbersmyr, Kjell Gunnar; Rogovchenko, Svitlana & Vysochinskiy, Dmitry (2021). An overview of Data Based Predictive Modeling Techniques used in Analysis of Vehicle Crash Severity. Fulltekst i vitenarkiv
• Kandukuri, Surya Teja & Robbersmyr, Kjell Gunnar (2021). Analytics for asset integrity management in wind farms (AIMWind). Fulltekst i vitenarkiv
• Jakobsen, Mette Mo; Tørdal, Sondre Sanden; Klausen, Andreas; Robbersmyr, Kjell Gunnar & Tyapin, Ilya (2021). Product development and project-based learning in Mechatronics, in the context of digitization and sustainability. Fulltekst i vitenarkiv
• Wæhle, Andreas; Nilsen, Jørgen; Tollevik, Sigurd; Rogne, Nora; Westbye, Stian & Robbersmyr, Kjell Gunnar (2021). The culture of product development in student organizations. Fulltekst i vitenarkiv
• Wæhle, Andreas Otter; Rogne, Nora; Westbye, Stian; Nilsen, Jørgen; Tollevik, Sigurd & Robbersmyr, Kjell Gunnar (2021). How to succeed with volunteers on the basis of commitment and ownership. Fulltekst i vitenarkiv
• Nyberg, Svein Olav Glesaaen; Robbersmyr, Kjell Gunnar; Holm, Jan Andreas & Sanfilippo, Filippo (2021). Static Tensile Experiments on Adhesion between Aluminium Profiles and Glass. Fulltekst i vitenarkiv Vis sammendrag

  In this work, we study the effects of the adhesion between aluminium profiles and glass from a static tensile perspective. A series of stretch curves from a 2016 experiment are analysed from their derivatives to find their points of float and maximum load bearing. The variable factors in this experiment are glass type, and type of connection: edge gluing, side (fugue) gluing, and excessive fugue gluing, which we have named fugue-edge. The stretch data gives us four quantities to analyse and compare: displacement and load at respectively float and at max load. We first display the results by simply comparing factor group against factor group. Whit this tool, only a few significant conclusions may be found. We then look at the more powerful statistical tools of linear regression and Anova (analysis of variance), explain these tools, and then draw conclusions about which factors are significant, and then about the size of the effect on the four variables under study. This forms the basis for our recommendation to window producers on where to focus their attention for the strongest possible glass-frame system.

• Knausgård, Kristian Muri; Skar, Siv; Sanfilippo, Filippo; Buldenko, Albert; Lindheim, Henning & Lunde, Jakob [Vis alle 8 forfattere av denne artikkelen] (2021). Branch-Manoeuvring Capable Pipe Cleaning Robot for Aquaponic Systems. Fulltekst i vitenarkiv Vis sammendrag

  Aquaponic systems are engineered ecosystems combining aquaculture and plant production. Nutrient rich water is continuously circulating through the system from aquaculture tanks. A biofilter with nitrifying bacteria breaks down fish metabolism ammonia into nitrite and nitrate, which plants and makes the aquaculture wastewater into valued organic fertiliser for the plants, contenting essential macro and micro elements. At the same time, plants are cleaning the water by absorbing ammonia from the fish tanks before it reaches dangerous levels for the aquatic animals. In principle, the only external input is energy, mainly in the form of light and heat, but fish food is also commonly provided. Growing fish food is potentially feasible in a closed loop system, hence aquaponic systems can possibly be an important source of proteins and other important nutrition when, for example, colonising other planets in the future. Fully autonomous aquaponic systems are currently not available. This work aims at minimising manual labour related to cleaning pipes for water transport. The cleaning process must be friendly to both plants and aquatic animals. Hence, in this work, pure mechanical cleaning is adopted. A novel belt-driven continuum robot capable of travelling through small/medium diameter pipes and manoeuvring branches and bends, is designed and tested. The robot is modular and can be extended with different cleaning modules through an interface providing CAN-bus network and electric power. The flexible continuum modules of the robot are characterised. Experimental results demonstrate that the robot is able to travel through pipes of diameter 50 mm to 75 mm, and also capable of handling T-branches of up to 90 degrees.

• Attestog, Sveinung; Huynh, Van Khang & Robbersmyr, Kjell Gunnar (2020). Improved Quadratic Time-Frequency Distributions for Fault Detection in Transient States. Fulltekst i vitenarkiv Vis sammendrag

  This paper aims to improve quadratic time-frequency distributions to adapt condition monitoring of electrical machines in transient states. Short-Time Fourier transform (STFT) has been a baseline signal processing technique for detecting fault characteristic frequencies. However, limits of window sizes due to loss of frequency- or time-resolution, make it hard to capture rapid changes in frequencies. Within this study, Choi-Williams and Wigner-Ville distributions are proposed to effectively detect peaks at characteristic frequencies while still maintaining low computation time. The improved quadratic time-frequency distributions allow for generating spectrograms of a longer lasting data signal and capturing multi-component signals with a better separation of the components than STFT. Further, the time resolution of the spectrograms generated by the proposed method is not affected by the window size. The effectiveness of the proposed methods is numerically verified from the data of an in-house test setup.

• Attestog, Sveinung; Huynh, Khang & Robbersmyr, Kjell Gunnar (2019). Modelling Demagnetized Permanent Magnet Synchronous Generator for Fault Diagnosis. Fulltekst i vitenarkiv Vis sammendrag

  Prediction of the lifetime for transformers is very important for maintenance and asset management. Finite element analysis was performed on a 5 MVA distribution transformers with aluminium foil-type windings and voltage rating 6600 V/23000 V. Electromagnetic modelling is implemented on the full three-phase transformer to calculate distributed losses, taking the skin effect into account. To reduce the computational burden, the distributed losses in one phase are used to analyse temperature rise in one phase of the transformer. The temperature rise results were used to determine the ageing rate of the transformer. Further, the influence of ambient temperature and cooling on the temperature rise and ageing rate of the transformer is numerically investigated.

• Senanayaka, Jagath Sri Lal; Huynh, Khang & Robbersmyr, Kjell Gunnar (2019). Electromagnetic and Vibration Modelling of an Electric Powertrain for a Simulation-driven Fault Diagnosis System. Fulltekst i vitenarkiv
• Davidsen, Jan Grimsrud; Hellum, Vidar; Robbersmyr, Kjell Gunnar; Jakobsen, Mette Mo; Sjursen, Aslak Darre & Frøiland, Eirik [Vis alle 7 forfattere av denne artikkelen] (2026). Automatiske sorteringssystemer for oppvandrende pukkellaks – evaluering av tre prototyper. NTNU Vitenskapsmuseet. ISSN 9788283224344. Fulltekst i vitenarkiv
• Noorsumar, Gulshan; Rogovchenko, Svitlana; Robbersmyr, Kjell Gunnar & Vysochinskiy, Dmitry (2023). Modeling, Simulation and Prediction of Vehicle Crashworthiness in Full Frontal Impact. Universitetet i Agder. ISSN 9788284271132. Fulltekst i vitenarkiv Vis sammendrag

  Vehicle crashworthiness assessment is critical to help reduce road accident fatalities and ensure safer vehicles for road users. Techniques to assess crashworthiness include physical tests and mathematical modeling and simulation of crash events, the latter is preferred as mathematical modeling is generally cheaper to perform in comparison with physical testing. The most common mathematical modeling technique used for crashworthiness assessment is nonlinear Finite Element (FE) modeling. However, a problem with the use of Finite Element Model (FEM) for crashworthiness assessment is inaccessibility to individual researchers, public bodies, small universities and engineering companies due to need for detailed CAD data, software licence costs along with high computational demands. This thesis investigates modeling strategies which are affordable, computationally and labour inexpensive, and could be used by the above-mentioned groups. Use of Lumped Parameter Models (LPM) capable of capturing vehicle parameters contributing to vehicle crashworthiness has been proposed as an alternative to adopting FEM, while the later have been used to validate LPMs developed in this thesis. The main crash scenario analysed is a full frontal impact against a rigid barrier. Front-end deformation which can be used to measure crash energy absorption and pitching which could lead to occupant injuries in a frontal crash event are parameters focused on. The thesis investigates two types of vehicles; vehicle with initial structure intact is defined as baseline vehicle, while a vehicle that underwent unprofessional repairs on its structural members made of Ultra High Strength Steel (UHSS) is defined as a modified vehicle. The proposed novel LPM for a baseline vehicle impact is inspired by pendulum motion and expresses the system using Lagrangian formulation to predict the two phases of impact: front-end deformation and vehicle pitching. Changes in crashworthiness performance of a modified vehicle were investigated with a FEM; tensile tests on UHSS coupons were conducted to generate material inputs for this FEM. Further, a full scale crash test was conducted to validate the FE simulations. An LPM to conduct crashworthiness assessment of a modified vehicle has been proposed, it is based on a double pendulum with a torsional spring representing the vehicle undergoing a full frontal impact.

• Attestog, Sveinung; Robbersmyr, Kjell Gunnar & Huynh, Khang (2022). Modelling and Detecting Faults of Permanent Magnet Synchronous Motors in Dynamic Operations. Universitetet i Agder. ISSN 9788284270951. Fulltekst i vitenarkiv Vis sammendrag

  Permanent magnet synchronous motors (PMSMs) have played a key role in commercial and industrial applications, i.e. electric vehicles and wind turbines. They are popular due to their high efficiency, control simplification and large torque-to-size ratio although they are expensive. A fault will eventually occur in an operating PMSM, either by improper maintenance or wear from thermal and mechanical stresses. The most frequent PMSM faults are bearing faults, short-circuit and eccentricity. PMSM may also suffer from demagnetisation, which is unique in permanent magnet machines. Condition monitoring or fault diagnosis schemes are necessary for detecting and identifying these faults early in their incipient state, e.g. partial demagnetisation and inter-turn short circuit. Successful fault classification will ensure safe operations, speed up the maintenance process and decrease unexpected downtime and cost. The research in recent years is drawn towards fault analysis under dynamic operating conditions, i.e. variable load and speed. Most of these techniques have focused on the use of voltage, current and torque, while magnetic flux density in the air-gap or the proximity of the motor has not yet been fully capitalised. This dissertation focuses on two main research topics in modelling and diagnosis of faulty PMSM in dynamic operations. The first problem is to decrease the computational burden of modelling and analysis techniques. The first contributions are new and faster methods for computing the permeance network model and quadratic time-frequency distributions. Reducing their computational burden makes them more attractive in analysis or fault diagnosis. The second contribution is to expand the model description of a simpler model. This can be achieved through a field reconstruction model with a magnet library and a description of both magnet defects and inter-turn short circuits. The second research topic is to simplify the installation and complexity of fault diagnosis schemes in PMSM. The aim is to reduce required sensors of fault diagnosis schemes, regardless of operation profiles. Conventional methods often rely on either steady-state or predefined operation profiles, e.g. start-up. A fault diagnosis scheme robust to any speed changes is desirable since a fault can be detected regardless of operations. The final contribution is the implementation of reinforcement learning in an active learning scheme to address the imbalance dataset problem. Samples from a faulty PMSM are often initially unavailable and expensive to acquire. Reinforcement learning with a weighted reward function might balance the dataset to enhance the trained fault classifier’s performance.

• Ali, Ahmed Mohamed Salem; Huynh, Khang & Robbersmyr, Kjell Gunnar (2021). Multilevel Inverters with Reduced Component Count for Energy Systems. Universitetet i Agder. ISSN 9788284270456. Fulltekst i vitenarkiv Vis sammendrag

  Multilevel inverters (MLIs) have gained increasing interest for advanced energy-conversion systems with a wide range of voltage levels, due to their attractive features of low harmonic contents, low dv/dt stress, low filtering requirements, low switching frequency, low electromagnetic interference (EMI), and employing low-rated semiconductor devices for producing high voltages. Further, some MLIs have a modularity feature, facilitating voltage and current scalability with high redundancy in switching states, allowing for fault-tolerant operations. The mentioned features are directly related to the number of generated voltage levels. However, enlarging level count renders challenges of requiring massive component counts, including DC sources, capacitors, power diodes, switches, inductors, and transformers. The high count of components negatively affects the size, cost, efficiency, lifespan, reliability, complexity of MLI-based energy conversion systems. Thus, proposing novel MLIs, which can enlarge the voltage level number with a low component count, is currently one of the most attractive topics in this research theme. In this dissertation, four three-phase topologies are proposed to mitigate the aforementioned shortcomings. Two transformerless topologies are proposed for low- and medium voltage applications, while two transformer-based topologies are intended for medium- and high-voltage applications. The proposed topologies have the key features of being capacitor-, diode-free, and low counts of DC sources, switches, and transformers. Further, three of them have a high modularity degree, allowing for higher voltage operations without increasing the voltage stress across the switches. The working principle of the proposed topologies are theoretically demonstrated, numerically verified, and experimentally validated through an in-house experimental setup. The effectiveness of the proposed topologies is proven through detailed comparative studies regarding component counts and voltage ratings. The proposed topologies are briefly described in this dissertation, while detailed explanations and results can be found in the appended papers.

• Senanayaka, Jagath Sri Lal & Robbersmyr, Kjell Gunnar (2020). Online Condition Monitoring of Electric Powertrains using Machine Learning and Data Fusion. Universitetet i Agder. ISSN 9788271179717. Fulltekst i vitenarkiv Vis sammendrag

  Safe and reliable operations of industrial machines are highly prioritized in industry. Typical industrial machines are complex systems, including electric motors, gearboxes and loads. A fault in critical industrial machines may lead to catastrophic failures, service interruptions and productivity losses, thus condition monitoring systems are necessary in such machines. The conventional condition monitoring or fault diagnosis systems using signal processing, time and frequency domain analysis of vibration or current signals are widely used in industry, requiring expensive and professional fault analysis team. Further, the traditional diagnosis methods mainly focus on single components in steady-state operations. Under dynamic operating conditions, the measured quantities are non-stationary, thus those methods cannot provide reliable diagnosis results for complex gearbox based powertrains, especially in multiple fault contexts. In this dissertation, four main research topics or problems in condition monitoring of gearboxes and powertrains have been identified, and novel solutions are provided based on data-driven approach. The first research problem focuses on bearing fault diagnosis at early stages and dynamic working conditions. The second problem is to increase the robustness of gearbox mixed fault diagnosis under noise conditions. Mixed fault diagnosis in variable speeds and loads has been considered as third problem. Finally, the limitation of labelled training or historical failure data in industry is identified as the main challenge for implementing data-driven algorithms. To address mentioned problems, this study aims to propose data-driven fault diagnosis schemes based on order tracking, unsupervised and supervised machine learning, and data fusion. All the proposed fault diagnosis schemes are tested with experimental data, and key features of the proposed solutions are highlighted with comparative studies.

• Hemmer, Martin & Robbersmyr, Kjell Gunnar (2020). Condition Monitoring Methods for Large, Low-speed Bearings. Universitetet i Agder. ISSN 9788271179601. Fulltekst i vitenarkiv Vis sammendrag

  In all industrial production plants, well-functioning machines and systems are required for sustained and safe operation. However, asset performance degrades over time and may lead to reduced effiency, poor product quality, secondary damage to other assets or even complete failure and unplanned downtime of critical systems. Besides the potential safety hazards from machine failure, the economic consequences are large, particularly in offshore applications where repairs are difficult. This thesis focuses on large, low-speed rolling element bearings, concretized by the main swivel bearing of an offshore drilling machine. Surveys have shown that bearing failure in drilling machines is a major cause of rig downtime. Bearings have a finite lifetime, which can be estimated using formulas supplied by the bearing manufacturer. Premature failure may still occur as a result of irregularities in operating conditions and use, lubrication, mounting, contamination, or external environmental factors. On the contrary, a bearing may also exceed the expected lifetime. Compared to smaller bearings, historical failure data from large, low-speed machinery is rare. Due to the high cost of maintenance and repairs, the preferred maintenance arrangement is often condition based. Vibration measurements with accelerometers is the most common data acquisition technique. However, vibration based condition monitoring of large, low-speed bearings is challenging, due to non-stationary operating conditions, low kinetic energy and increased distance from fault to transducer. On the sensor side, this project has also investigated the usage of acoustic emission sensors for condition monitoring purposes. Roller end damage is identified as a failure mode of interest in tapered axial bearings. Early stage abrasive wear has been observed on bearings in drilling machines. The failure mode is currently only detectable upon visual inspection and potentially through wear debris in the bearing lubricant. In this thesis, multiple machine learning algorithms are developed and applied to handle the challenges of fault detection in large, low-speed bearings with little or no historical data and unknown fault signatures. The feasibility of transfer learning is demonstrated, as an approach to speed up implementation of automated fault detection systems when historical failure data is available. Variational autoencoders are proposed as a method for unsupervised dimensionality reduction and feature extraction, being useful for obtaining a health indicator with a statistical anomaly detection threshold. Data is collected from numerous experiments throughout the project. Most notably, a test was performed on a real offshore drilling machine with roller end wear in the bearing. To replicate this failure mode and aid development of condition monitoring methods, an axial bearing test rig has been designed and built as a part of the project. An overview of all experiments, methods and results are given in the thesis, with details covered in the appended papers.

• Munyazikwiye, Bernard & Robbersmyr, Kjell Gunnar (2020). Mathematical Modelling and Analysis of Vehicle Frontal Crash using Lumped Parameters Models. Universitetet i Agder. ISSN 9788271179618. Fulltekst i vitenarkiv Vis sammendrag

  A full-scale crash test is conventionally used for vehicle crashworthiness analysis. However, this approach is expensive and time-consuming. Vehicle crash reconstructions using different numerical modelling approaches can predict vehicle behavior and reduce the need for multiple full-scale crash tests, thus research on the crash reconstruction has received a great attention in the last few decades. Among modelling approaches, lumped parameters models (LPM) and finite element models (FEM) are commonly used in the vehicle crash reconstruction. This thesis focuses on developing and improving the LPM for vehicle frontal crash analysis. The study aims at reconstructing crash scenarios for vehicle-to-barrier (VTB), vehicleoccupant (V-Occ), and vehicle-to-vehicle (VTV), respectively. In this study, a single mass-spring-damper (MSD) is used to simulate a vehicle to-barrier or a wall. A double MSD is used to model the response of the chassis and passenger compartment in a frontal crash, a vehicle-occupant, and a vehicle-tovehicle, respectively. A curve fitting, state-space, and genetic algorithm are used to estimate parameters of the model for reconstructing the vehicle crash kinematics. Further, the piecewise LPM is developed to mimic the crash characteristics for VTB, VO, and VTV crash scenarios, and its predictive capability is compared with the explicit FEM. Within the framework, the advantages of the proposed methods are explained in detail, and suggested solutions are presented to address the limitations in the study.

• Klausen, Andreas & Robbersmyr, Kjell Gunnar (2019). Condition monitoring of rolling element bearings during low and variable speed conditions. Universitetet i Agder. ISSN 9788271179267. Fulltekst i vitenarkiv

Se alle arbeider i NVA