Research on fault identification of power equipment based on bp neural network

Since the tenth five-year plan, grid-connected wind power has been developing rapidly in Inner Mongolia, Hebei, and Liaoning, easing local residents’ pressure to use electricity. But the frequent failure of the gearbox of wind turbines restricts the development of wind power construction. Based on this, this study applied Back Propagation (BP) neural network technology applied to the diagnosis of wind turbine gearbox faults. The number of nodes in each layer was determined. The simulation was carried out using the MATLAB software. The results showed that the identification accuracy reached 96.5% and 95.2% respectively in wear fault gears and rolling bearing fault identification tests. The identification results verify the feasibility of the BP neural network for fault identification of power equipment. This study provides a guarantee for solving the problem of gearbox failure in wind turbines. © 2021, Cefin Publishing House. All rights reserved.