Vibration Fault Diagnosis of Circuit Breaker Based on CGWO-VMD and ELM Combined with PCA

In order to effectively extract the characteristic information of circuit breaker vibration signal and improve the classification accuracy, this paper proposes a circuit breaker fault state diagnosis method based on chaotic gray wolf optimizer algorithm (CGWO) combined with variational mode decomposition (VMD). The algorithm (CGWO) optimizes Variational Mode Decomposition (VMD) and uses extreme value learning machine (ELM) in combination with principal component analysis (PCA) to complete fault classification. First, since the GWO optimization process is prone to fall into local optimum and has slow convergence speed, a sinusoidal chaotic map is introduced to improve GWO, and CGWO is used to optimize VMD to obtain the optimal number of modes K and quadratic penalty factor. Secondly, the principal component analysis of the modal components is carried out to form the fault eigenvectors. Finally, the feature vectors are input into the ELM model to classify and identify different types of faults. Through experimental analysis, compared with GWO-ELM and HBA-ELM, this method has a better diagnosis effect on composite faults, and the fault classification accuracy rate reaches to almost 100%, which provides a new way to solve the problem of circuit breaker fault diagnosis.