TY - JOUR
T1 - Intensive Multi-order Feature Extraction for Incipient Fault Detection of Inverter System
AU - Wang, Min
AU - Cheng, Feiyang
AU - Xie, Min
AU - Qiu, Gen
AU - Zhang, Jingxin
PY - 2025/2
Y1 - 2025/2
N2 - Inverter systems play a crucial role in aerospace, defense, transportation, modern industry, and power systems, leading to extensive efforts from scholars and engineers in fault diagnosis. Data-based methods are widely utilized with the accessible history data instead of complex math modeling for this issue but they are incompetent for obstinate incipient fault. Therefore, this paper proposes an intensive multi-order feature extractor (IMFE) for the incipient fault detection of inverter system, with intensively extracting deep statistical features and reducing harmful perturbations. Firstly, a dense structure with short paths between non-adjacent layers is adopted for multi-order knowledge re-utilization. Then, the acquired features are refined and the low-quality information is discarded. In addition, the effectiveness of IMFE is demonstrated through rigorous mathematical derivation with sensitivity and complexity analysis. Finally, a three-phase inverter system platform based on current and voltage dual control is established to verify the superiority of the proposed method. Experimental results show that the proposed approach significantly improves fault detection performance, achieving a 3.1 % higher fault detection rate compared to existing state-of-the-art methods. © 2024 IEEE.
AB - Inverter systems play a crucial role in aerospace, defense, transportation, modern industry, and power systems, leading to extensive efforts from scholars and engineers in fault diagnosis. Data-based methods are widely utilized with the accessible history data instead of complex math modeling for this issue but they are incompetent for obstinate incipient fault. Therefore, this paper proposes an intensive multi-order feature extractor (IMFE) for the incipient fault detection of inverter system, with intensively extracting deep statistical features and reducing harmful perturbations. Firstly, a dense structure with short paths between non-adjacent layers is adopted for multi-order knowledge re-utilization. Then, the acquired features are refined and the low-quality information is discarded. In addition, the effectiveness of IMFE is demonstrated through rigorous mathematical derivation with sensitivity and complexity analysis. Finally, a three-phase inverter system platform based on current and voltage dual control is established to verify the superiority of the proposed method. Experimental results show that the proposed approach significantly improves fault detection performance, achieving a 3.1 % higher fault detection rate compared to existing state-of-the-art methods. © 2024 IEEE.
KW - Fault detection
KW - Feature ensemble
KW - Inverter system
KW - Reliability evaluation
UR - http://www.scopus.com/inward/record.url?scp=85208384602&partnerID=8YFLogxK
UR - https://www.scopus.com/record/pubmetrics.uri?eid=2-s2.0-85208384602&origin=recordpage
U2 - 10.1109/TPEL.2024.3487853
DO - 10.1109/TPEL.2024.3487853
M3 - RGC 21 - Publication in refereed journal
SN - 0885-8993
VL - 40
SP - 3543
EP - 3552
JO - IEEE Transactions on Power Electronics
JF - IEEE Transactions on Power Electronics
IS - 2
ER -