GIS Fault Prediction Approach Based on IPSO-LSSVM Algorithm

With the improvement of industrialization, importance equipment failure prediction is increasing day by day. Accurate gas-insulated switchgear (GIS) in advance can reduce economic loss caused power system to operate normally. Therefore, a GIS fault approach based on Improved Particle Swarm Optimization Algorithm (IPSO)-least squares support vector machine (LSSVM) proposed this paper. Firstly, future gas conditions determine characteristic data SF6 decomposition are analyzed; Secondly, model LSSVM established, and IPSO algorithm used normalize parameters LSSVM. The c radial basis kernel function σ2 optimized, which meet needs later search accuracy while ensuring global capability early stage. Finally, effectiveness method verified switch. Simulation results shows that, compared with methods IGA-LSSVM PSO-LSSVM, rate reached 92.1%, has smallest absolute error, higher stronger ability.