MPCA-based train suspension system fault analysis method and system

Abstract

The invention discloses an MPCA-based train suspension system fault analysis method. According to the method disclosed in the invention, a multilinear principal component analysis method (MPCA) is applied to the fault diagnosis of railway vehicle suspension systems. In order to find the weak faults of the suspension systems conveniently and acquire fault information as much as possible, acquired original two-dimensional data is constructed into a third-order tensor form, the advantages, of processing the tensor data, of the MPCA is utilized to decrease the variable and temporal correlation in a local neighborhood as far as possible, and dimension reduction processing and characteristic extraction are carried out on training samples (regarded as tensor objects) in a plurality of (mode) directions, so that the structure and dependency of the original data are protected. Each sample is expressed by using an information amount which is least and has most remarkable characteristics to the greatest extent, so that the transformed low-dimensionality sub-spaces have good mode expression ability, and the calculation amount is reduced. According to the method disclosed in the invention, the train suspension system weak fault detection ability can be greatly enhanced, and the safety performance of the train traveling is improved.

Description

technical field [0001] The invention relates to the field of train fault analysis, in particular to a train suspension system fault analysis method and system based on a multi-linear principal component analysis (MPCA) algorithm. Background technique [0002] With the rapid development of my country's urban rail transit, issues such as the safety and reliability of rail transit vehicle systems have attracted more and more attention. For big cities, subway transportation is undoubtedly the most important part of urban public transportation. However, especially in the morning and evening peak periods, the subway vehicles are overloaded for a long time, which makes the performance of the suspension components of the subway vehicles gradually decay after the vehicles are put into operation, and even sudden failures may occur. The actual maintenance experience of the subway operation company shows that after the subway vehicle is put into operation for one to two years, some sus...

AI Technical Summary

Problems solved by technology

[0005] The technical problem to be solved by the present invention is to provide a train suspension system fault analysis method and system based on the multi-linear principal component analysis (MPCA) algorithm, so as to solve the problem that the existing train suspension system fault diagnosis process is limited to the exact knowledge of model parameters The model-based method is used to detect or estimate the faults of the system. The data-driven method is still unable to detect weak faults. It is not sensitive enough to the performance attenuation of the suspension system components (mainly springs and dampers), and it is difficult to eliminate weak faults. Hidden dangers and other issues

Images