Title: An improved extension neural network methodology for fault diagnosis of complex electromechanical system

Authors: Yunfei Zhou; Yunxiu Sai; Li Yan

Addresses: School of Management, Xi'an University of Finance and Economics, Xi'an 710100, China ' Xi'an Shiyou University, Xi'an 710065, China ' Xi'an Technological University, Xi'an 710021, China

Abstract: Fault diagnosis of complex electromechanical system is always a complex and challenging problem. The method of how to identify, extract and classify fault features is still a key problem. Towards the end, this study proposes an intelligent fault diagnosis method based on extension neural network and uniform distribution search for particle swarm optimisation algorithm. Specifically, the influence of joint field on the classical domain is considered in the training dataset and the improved particle swarm optimisation algorithm is used to optimise the classical domain. Subsequently, the improved extended distance is used for classification training. The simulation results of turbine generator set prove that the method cannot only correctly extract the available classical domain features from the collected training dataset, but also has higher diagnostic accuracy and fast training process.

Keywords: fault diagnosis; complex electromechanical system; CES; particle swarm optimisation; PSO; extension neural network; ENN.

DOI: 10.1504/IJBIC.2021.119950

International Journal of Bio-Inspired Computation, 2021 Vol.18 No.4, pp.250 - 258

Received: 28 Apr 2020
Accepted: 27 Sep 2020
Published online: 04 Jan 2022 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article