Rolling Bearing Fault Diagnosis Method Using Glowworm Swarm Optimization and Artificial Neural Network

Fault diagnosis has long been recognised as one of the most effective methods of reducing operation and maintenance cost in rotating industry, especially in bearings. A method based on BP neural network modified by glowworm swarm optimization (GSO) was proposed for fault diagnosis of rolling bearings. Six fault features were selected as the input of network. GSO algorithm was applied to simultaneously optimize the initial weight and threshold values of BP neural network. The reliability of the proposed technique was confirmed by experimental data, which indicated the potential applications of this method in the field of rolling bearing fault diagnosis.