Electric locomotive fault prediction and health management method

Abstract

The invention discloses an electric locomotive fault prediction and health management method. Sensors arranged in and out of an electric locomotive are used for collecting and monitoring state characteristic parameter data of different components of the electric locomotive and external environmental parameter data in real time; the parameter data is arranged, compressed and coded and sent to a ground processing center through a wireless network; the ground processing center receives the data transmitted from the electric locomotive and establishes state evaluation models for the different components of the electric locomotive through fuzzy logic to analyze corresponding data obtained after extraction is carried out according to time sequence classification, and by means of estimation, the performance, states and health situations of the different components of the electric locomotive are obtained; a fault prediction model is established based on an artificial neural network through a fault foreboding and failure mechanism diagnosing and predicting method, fault prediction is carried out, the residual life is calculated, and data is recorded to form a historical database. By means of the electric locomotive fault prediction and health management method, fault prediction can be automatically carried out, the residual life of the components can be calculated, and health management of the electric locomotive can be achieved.

Description

technical field [0001] The invention relates to the technical field of electric locomotives, in particular to a method for fault prediction and health management of electric locomotives. Background technique [0002] The traditional maintenance methods of electric locomotives include preventive maintenance and breakdown maintenance. Preventive maintenance is time-based periodic maintenance. The maintenance cycle determined by this method is conservative and prone to transitional maintenance. Fault maintenance is the maintenance after the equipment fails, and this method can easily lead to significant economic losses of the equipment. [0003] Because the traditional electric locomotive maintenance method has the problems of long maintenance period and high cost, a new maintenance method emerges, that is, condition-based maintenance. Condition-based maintenance monitors the operation of equipment and predicts failures. On this basis, the remaining life estimation, maintenan...

AI Technical Summary

Problems solved by technology

[0004] At present, the condition-based maintenance monitors the operation of equipment mostly through artificial analysis of monitoring data and artificial prediction of faults. When the detection data is large, it is difficult for humans to accurately judge

Moreover, the monitoring data of each electric locomotive is relatively scattered, and it is difficult to centrally judge the operation status of each electric locomotive in a timely manner