Research On Multi-signal Model Fault Diagnosis Method Considering Test Uncertainty

With the improvement of the function and performance of electronic equipment,the degree of integration is getting higher and higher,and the complexity of the circuit is also increasing,which puts forward higher requirements for the maintenance level of maintenance personnel.If the maintenance of the circuit system and how to detect when the fault occurs are not properly designed at the beginning of the system design,it will take a long time to repair the fault.Many scholars have done a lot of research on circuit fault diagnosis,but there are relatively few studies that consider the test uncertainty,but the introduction of the test point test uncertainty is necessary.When the test uncertainty is not considered.It is a test under ideal conditions and does not conform to objective laws.Therefore,it is of great significance to study the circuit fault diagnosis method considering the test uncertainty.Based on the actual engineering project of "Development of XXX Equipment Fault Diagnosis System" and based on the multisignal model,thesis studies the measurement point optimization strategy and fault diagnosis considering test uncertainty in circuit testing.The main research contents of thesis are as follows :(1)Modeling and analysis of the multi-signal model considering the test uncertainty.Firstly,the multi-signal model under the test uncertainty of the circuit measuring point is constructed,and then the method of determining the specific value in the D matrix is ??introduced and analyzed.Finally,the fault detection rate,isolation rate,false alarm rate,missed detection rate,and test cost under the model are presented,and the constraint conditions of detection rate and isolation rate are constructed,and the false alarm rate,missed detection rate and Mathematical description of multi-objective constraints with test cost as objective.(2)Research on the optimal selection method of measuring points under test uncertainty.Aiming at the problem of high false alarm rate and high missed detection rate in circuit fault diagnosis,thesis uses the Binary Multi-Objective Particle Swarm Optimization Algorithm Base on Crowding Distance(BMOPSO-CD algorithm)based on crowding distance to Optimal selection of test points is carried out to reduce the false alarm rate,missed detection rate and test cost as much as possible while ensuring the detection rate and isolation rate.The method is verified with the actual data,and compared with the NSGA-? algorithm,the performance of this method is better.(3)Research on fault location method under test uncertainty.Thesis regards this problem as a non-integer weighted set covering problem,which is a NP-complete problem.In thesis,an approximate algorithm based on the greedy algorithm idea is used to solve the problem,and finally the solutions of single-fault and multi-fault problems are given.Finally,the superiority of this method is verified by a practical case.(4)For the power supply module of a circuit board in the actual project,firstly perform multi-signal modeling under test uncertainty,including circuit module division,fault and measurement point settings;then use the multi-objective particle algorithm to the circuit board.The optimal selection of measurement points reduces the maintenance and testing costs while ensuring that the indicators meet the project requirements.Finally,the method in this paper is used to isolate the faults,and the final isolation rate meets the project requirements.