Sequential-test dynamic adjustment method based on AO* algorithm

Abstract

The invention discloses a sequential-test dynamic adjustment method based on the AO* algorithm. The sequential-test dynamic adjustment method includes the steps that an optimal fault diagnosis tree of an electronic system is obtained through the AO* algorithm; then the testing cost of a test and the probability dynamic condition of a fault are monitored, and when the testing cost of the test is increased, a fault node selecting a test point is reevaluated; when the testing cost of the test is reduced, a fault node which does not select the test point is reevaluated, and the optimal fault diagnosis tree is adjusted; when the probability of the fault is changed, the optimal fault diagnosis tree is adjusted. According to changes of the testing cost and changes of the fault probability in the electronic system, the present optimal fault diagnosis tree is correspondingly adjusted, the new optimal fault diagnosis tree is rapidly obtained, and adaptive dynamic adjustment of the sequential test is achieved.

Description

technical field [0001] The invention belongs to the technical field of electronic system fault diagnosis, and more specifically relates to a sequential test dynamic adjustment method based on the AO* algorithm. Background technique [0002] In electronic system fault diagnosis technology, the sequential testing problem is defined as a five-tuple problem (S, P, T, C, D). where, S={s 0 ,s 1 ,s 2 ,...,s M} represents a finite set of system fault states, where s 0 Indicates the state of the system without faults, s 1 to s M Indicates the state of different faults in the system. P={p 0 ,p 1 ,p 2 ,...,p M} is the prior failure probability vector of occurrence of each system state. Assuming that the system can only be in a certain fault state or no fault state, it is necessary to normalize the prior fault probability vector P. T={t 1 ,t 2 ,...,t N} are N available test sets of the system, C={c 1 ,c 2 ,...,c N} represents the corresponding test cost vector, where ...

AI Technical Summary

Problems solved by technology

[0012] The current sequential testing methods are all static, that is, all parameters are static during the calculation of sequential testing, but in actual engineering applications, the parameters are likely to change

In the existing algorithm, when the parameters change, the AO* algorithm needs to be called again to generate a new optimal fault diagnosis tree. When the scale of the electronic system is large, the number of fault states and the number of measurement points are relatively large. If the parameter changes each time The optimal fault diagnosis tree is regenerated every time, obviously there is a problem of low efficiency

Images