Switchgear Fault Diagnosis Based on Support Vector Machine Optimized by an Improved Dung Beetle Optimization Algorithm

Power switchgear, as one of the important electrical equipment in the distribution network, plays a crucial role to play in ensuring the reliability, safety and continuity of the distribution network with its stable operation. To promptly detect the operational state of switchgear, this article proposes a fault diagnosis model based on IDBO-SVM. Twelve sets of state data during the operation of four areas of the switchgear are collected to create the switchgear fault diagnosis dataset. The dung beetle population is initialized using a chaotic model, and the dung beetle optimization algorithm is enhanced through Cauchy's variation and adaptive weight adjustment strategies to enhance the algorithm's parameter optimization capability for support vector machines. The analysis of the case shows that the overall detection accuracy of the model for the switchgear operation state is 98.28%, which indicates that the IDBO-SVM fault detection model can effectively achieve precise identification of switchgear fault operation state, and can provide technical support for the technicians to carry out the overhaul and maintenance of the switchgear.