Pavan Kumar Mp
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Pavan Kumar Mp has a total of 4 co-authors. Here are the top co-authors by number of shared publications:
- Kun-Chih Chen
Institute of Electronics, National Yang Ming Chiao Tung University, Hsinchu, Taiwan
3 shared publications - Hsu-Chi Chen
Institute of Electronics, National Yang Ming Chiao Tung University, Hsinchu, Taiwan
2 shared publications - Zhe-Xiang Tu
Department of Computer Science and Engineering, National Sun Yat-Sen University, Kaohsiung, Taiwan
2 shared publications Zi-Jie Gao
Department of Computer Science and Engineering, National Sun Yat-Sen University, Kaohsiung, Taiwan
1 shared publication
Latest Publications
- 14 October 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Instrumentation and Measurement
- Vol. 73 (00189456) , 1-13
Enhancing Learning in Fine-Tuned Transfer Learning for Rotating Machinery via Negative Transfer Mitigation
Rotating machinery plays a crucial role in industries such as petroleum, automotive, and food processing, relying on bearings for smooth movement. However, the wear and degradation of bearings due to fluctuating speeds, excessive loads, and prolonged operation pose significant challenges. Therefore, highlighting the essential role of real-time monitoring and diagnostics is critical for preventing machinery failures, improving equipment reliability, and reducing maintenance costs. While Predictive Maintenance (PdM) is widely employed for machinery health monitoring using sensor data, the analysis of diverse and voluminous data poses notable difficulties. To address this problem, we first propose a Normalized Exponential Power Entropy (NEPE)-based feature extraction method, which facilitates fault diagnosis when combined with a Neural Network (NN). Integrating Deep Learning (DL) into Prognostics and Health Management (PHM) improves fault diagnosis but faces limited labeled data and generalizability issues. Although Transfer Learning (TL) overcomes these limitations, fine-tuning with limited target domain data can lead to Negative Transfer Learning (NTL) and Catastrophic Forgetting (CF). To address this problem, we introduce Dynamic-Feature Adaptive Thresholded Normalization (D-FATN), a novel regularization approach that selectively modulates normalized input features during fine-tuning. D-FATN prevents NTL and mitigates CF, improving fault diagnosis. Experimental results demonstrate that our proposed method can improve 7.35% to 30.55% accuracy compared with other feature extraction methods. Additionally, our proposed D-FATN further improves 1.4% to 11.7% accuracy under identical and varying working conditions across both public and real measurement datasets.
- 14 October 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Instrumentation and Measurement
- Vol. 73 (00189456) , 1-13
Enhancing Learning in Fine-Tuned Transfer Learning for Rotating Machinery via Negative Transfer Mitigation
Rotating machinery plays a crucial role in industries such as petroleum, automotive, and food processing, relying on bearings for smooth movement. However, the wear and degradation of bearings due to fluctuating speeds, excessive loads, and prolonged operation pose significant challenges. Therefore, highlighting the essential role of real-time monitoring and diagnostics is critical for preventing machinery failures, improving equipment reliability, and reducing maintenance costs. While Predictive Maintenance (PdM) is widely employed for machinery health monitoring using sensor data, the analysis of diverse and voluminous data poses notable difficulties. To address this problem, we first propose a Normalized Exponential Power Entropy (NEPE)-based feature extraction method, which facilitates fault diagnosis when combined with a Neural Network (NN). Integrating Deep Learning (DL) into Prognostics and Health Management (PHM) improves fault diagnosis but faces limited labeled data and generalizability issues. Although Transfer Learning (TL) overcomes these limitations, fine-tuning with limited target domain data can lead to Negative Transfer Learning (NTL) and Catastrophic Forgetting (CF). To address this problem, we introduce Dynamic-Feature Adaptive Thresholded Normalization (D-FATN), a novel regularization approach that selectively modulates normalized input features during fine-tuning. D-FATN prevents NTL and mitigates CF, improving fault diagnosis. Experimental results demonstrate that our proposed method can improve 7.35% to 30.55% accuracy compared with other feature extraction methods. Additionally, our proposed D-FATN further improves 1.4% to 11.7% accuracy under identical and varying working conditions across both public and real measurement datasets.
- 9 October 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Instrumentation and Measurement
- Vol. 73 (00189456) , 1-16
Mitigating Negative Transfer Learning in Source Free-Unsupervised Domain Adaptation for Rotating Machinery Fault Diagnosis
Driven by the data-centric industrial revolution, Prognostics and Health Management (PHM) technology based on sophisticated Deep Learning (DL) methods has become crucial for industrial fault diagnosis. Based on well-labeled data, the DL methods further improve PHM development. However, the efficiency of the involved DL methods is impacted significantly due to the limited data collection. To address this problem, Source Free-Unsupervised Domain Adaptation (SF-UDA) has been proposed to improve DL efficiency without fully labeled source data. However, the traditional SF-UDA methods are constrained by predefined label numbers and insufficient distribution alignment. To solve this problem, we propose the use of Initial Pseudo Labels (IPLs) generated through Adaptive Bandwidth-based Mean Shift Clustering (ABMSC), which are then refined using cosine similarity with unlabeled target data, bolstering SF-UDA’s efficiency. Furthermore, we introduce Weight-Aware Regularization (W-AR) to effectively mitigate negative transfer. Experimental results show significant accuracy improvements, with gains of 1.5%-6.7% and 1.3%-5% in the same operating environment, and 0.4%-24.8% and 3.1%-8.6% in different operating environments.
- 9 October 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Instrumentation and Measurement
- Vol. 73 (00189456) , 1-16
Mitigating Negative Transfer Learning in Source Free-Unsupervised Domain Adaptation for Rotating Machinery Fault Diagnosis
Driven by the data-centric industrial revolution, Prognostics and Health Management (PHM) technology based on sophisticated Deep Learning (DL) methods has become crucial for industrial fault diagnosis. Based on well-labeled data, the DL methods further improve PHM development. However, the efficiency of the involved DL methods is impacted significantly due to the limited data collection. To address this problem, Source Free-Unsupervised Domain Adaptation (SF-UDA) has been proposed to improve DL efficiency without fully labeled source data. However, the traditional SF-UDA methods are constrained by predefined label numbers and insufficient distribution alignment. To solve this problem, we propose the use of Initial Pseudo Labels (IPLs) generated through Adaptive Bandwidth-based Mean Shift Clustering (ABMSC), which are then refined using cosine similarity with unlabeled target data, bolstering SF-UDA’s efficiency. Furthermore, we introduce Weight-Aware Regularization (W-AR) to effectively mitigate negative transfer. Experimental results show significant accuracy improvements, with gains of 1.5%-6.7% and 1.3%-5% in the same operating environment, and 0.4%-24.8% and 3.1%-8.6% in different operating environments.
- 24 February 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal on Emerging and Selected Topics in Circuits and Systems
- Vol. 13 (2) , 514-523
Time Series-Based Sensor Selection and Lightweight Neural Architecture Search for RUL Estimation in Future Industry 4.0
Industry 4.0 is the current development trend in manufacturing technology. Regarding prognostic and health management (PHM) systems, a major task is to predict the remaining useful life (RUL) of a machine. To have precise RUL information, massive sensors should be involved, which leads to the dramatic cost of sensor network construction. Therefore, many sensor selection methods have been proposed to remove the redundant sensors under a certain constraint of RUL estimation. However, those approaches do not consider time-series sensing data, which is an intrinsic feature of mechanical signals. Therefore, the current research suffers from sensor under-killing problems due to the pessimistic consideration. To solve this problem, this work first considers the time-series sensing data to propose an integrated group-based valuable sensor selection algorithm by using the Least Absolute Shrinkage and Selection Operator (LASSO) property. In addition, to reduce the computing and hardware overhead, we further propose a heuristic neural architecture search (NAS) method to reduce the number of neurons in each neural network for further RUL estimation. To verify the proposed approaches, we use the Commercial Modular Aero-Propulsion System Simulation (C-MAPSS) dataset and utilize the PHM score to evaluate the performance of RUL estimation. Compared with the conventional methods, the proposed sensor selection method can reduce the PHM score by 9% to 99% with fewer sensors. On the other hand, the proposed NAS method can reduce the number of neurons by 86% to 89% and reduce the PHM score by 17% to 79% compared with the traditional NAS approaches.
- 24 February 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal on Emerging and Selected Topics in Circuits and Systems
- Vol. 13 (2) , 514-523
Time Series-Based Sensor Selection and Lightweight Neural Architecture Search for RUL Estimation in Future Industry 4.0
Industry 4.0 is the current development trend in manufacturing technology. Regarding prognostic and health management (PHM) systems, a major task is to predict the remaining useful life (RUL) of a machine. To have precise RUL information, massive sensors should be involved, which leads to the dramatic cost of sensor network construction. Therefore, many sensor selection methods have been proposed to remove the redundant sensors under a certain constraint of RUL estimation. However, those approaches do not consider time-series sensing data, which is an intrinsic feature of mechanical signals. Therefore, the current research suffers from sensor under-killing problems due to the pessimistic consideration. To solve this problem, this work first considers the time-series sensing data to propose an integrated group-based valuable sensor selection algorithm by using the Least Absolute Shrinkage and Selection Operator (LASSO) property. In addition, to reduce the computing and hardware overhead, we further propose a heuristic neural architecture search (NAS) method to reduce the number of neurons in each neural network for further RUL estimation. To verify the proposed approaches, we use the Commercial Modular Aero-Propulsion System Simulation (C-MAPSS) dataset and utilize the PHM score to evaluate the performance of RUL estimation. Compared with the conventional methods, the proposed sensor selection method can reduce the PHM score by 9% to 99% with fewer sensors. On the other hand, the proposed NAS method can reduce the number of neurons by 86% to 89% and reduce the PHM score by 17% to 79% compared with the traditional NAS approaches.