Abstract
The paper is devoted to current issues of locomotive asynchronous traction motor (ATEM) fault detection using neural networks. Developed sophisticated intelligent methods for monitoring and inspecting the technical condition of ATE bearings. Current absorption spectra are analysed to assess the technical condition of the induction bearing units. The mechanical vibration frequencies of a squirrel cage induction motor are presented. The method of artificial neural networks which are universal approximators and can effectively and efficiently solve problems of monitoring and diagnostics of technical condition of locomotive induction traction motors is applied. A neural network model and framework for monitoring the technical condition of ATED bearings has been developed. They are based on rules and user-provided facts to recognise the situation, make a diagnosis, formulate a solution or make a recommendation. The main failures of the bearing units of squirrel cage ATED are analysed. A methodology has been developed to build a neural network model of the ATED. The structure and architecture of the artificial neural network is defined. An experimental research has been conducted. The results enable the determination of bearing faults in asynchronous traction motors with squirrel cage rotor.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call