Research on Fault Diagnosis of Ship Power System Based on Improved Particle Swarm Optimization Neural Network Algorithm

Abstract

A fault diagnosis method for ship power system based on optimized particle swarm neural network is proposed. The optimization ability of particle swarm optimization is improved by particle dynamic optimization management and cross strategy. The threshold and weight of neural network are optimized by improved particle swarm optimization. The advantages of the algorithm is strong global search and fast convergence. It shows better speed and accuracy in the case of ship generator fault diagnosis.