TY  - EJOU
AU  - Wu, Zhao 
AU  - Jin, Ri 
AU  - Wang, Haoyu 
AU  - Zhang, Mingming 
AU  - Chen, Zhenfei 

TI  - Stator Single-Phase Ground Fault Diagnosis of Large Hydro-Generators Based on IDBO-Optimized VMD Parameters
T2  - Energy Engineering

PY  - 
VL  - 
IS  - 
SN  - 1546-0118

AB  - To address the challenge of diagnosing stator single-phase ground faults in large fractional-slot hydro-generators, this study proposes an Improved Dung Beetle Optimizer (IDBO)-based Variational Mode Decomposition (VMD) method. Since VMD is highly sensitive to the selection of the number of modes <i>K</i> and the penalty factor <i>α</i>, improper parameter settings may result in incomplete signal decomposition and loss of fault-related features. To overcome this limitation, the conventional Dung Beetle Optimizer (DBO) is enhanced by incorporating a Logistic–Tent hybrid chaotic initialization, a dynamic convergence factor, and a random perturbation mechanism, enabling more accurate optimization of VMD parameters. A stator single-phase ground fault model is constructed to generate fault signals, which are decomposed using the optimized VMD. Sample entropy features are then extracted from the intrinsic mode functions and fed into a Slime Mould Algorithm (SMA)-optimized Support Vector Machine (SVM) for classification. The proposed IDBO achieves faster convergence and lower envelope entropy compared with DBO. Experimental results show that the proposed method attains a diagnostic accuracy of 98.13%, demonstrating its effectiveness in single-phase stator ground fault diagnosis.
KW  - Fractional-slot hydro-generator; single-phase ground fault; improved dung beetle optimizer; variational mode decomposition; support vector machine

DO  - 10.32604/ee.2026.078619