Built-in test method and system for mechanical system

Abstract

The invention relates to a built-in test method and system for a mechanical system. The method comprises the steps that: a fault symptom space matrix corresponding to the mechanical system is built, each line of data of the matrix is used for representing that a downstream monitoring point of a fault component in the mechanical system outputs data abnormally and other monitoring points output data normally after the corresponding component fails, each column of the matrix represents a monitoring point, and different rows of the matrix correspond to different components of the mechanical system; after the mechanical system runs, the output data of all the monitoring points are obtained, a symptom monitoring data set is established according to the output data of all the monitoring points, and all data in the data set are used for representing whether the output data of the different monitoring points are abnormal or not; when abnormal data exists in the data set, fault diagnosis is carried out on the basis of the matrix and the symptom monitoring data set, and diagnosed fault types comprise at least one of a determined fault and an undetermined fault; and when the diagnosed faults comprise the undetermined fault, fault diagnosis until is continued until a certain fault is obtained.

Description

technical field [0001] This paper relates to testing technology, especially a method and system for in-machine testing of mechanical systems. Background technique [0002] The fast fault diagnosis of the mechanical system usually needs to be combined with the BIT (Built-In Test) design to realize it. In the traditional BIT design of mechanical systems, fault diagnosis is usually based on the independent test results of individual components, but because the components of the mechanical system are interconnected and coupled, the failure of a single component is usually reflected simultaneously on multiple different monitoring points , correspondingly, the monitoring data obtained from one monitoring point may also come from multiple different faulty components. Therefore, the fault source cannot be accurately judged only by the monitoring data of a single monitoring point. Contents of the invention [0003] The present application provides an in-machine testing method and...

AI Technical Summary

Problems solved by technology

Therefore, only relying on the monitoring data of a single monitoring point cannot accurately determine the source of the fault

Images