Research On Computer Aided Testability Analysis And Design Technology Of Electronic System

In order to ensure the stability and safety of all kinds of electronic equipment,especially for weapons and aerospace equipment,daily maintenance work is essential.However,not only the detection cost is very high,but also the detection is increasingly difficult,so it is urgent to enhance its testability.Through testability analysis of the system,it can optimize the arrangement of test points,achieve the purpose of improving diagnosis efficiency and reducing test cost.Therefore,this paper focuses on the auxiliary analysis and design of testability of electronic systems.Two new algorithms are proposed to solve the diagnosis strategy,and some improvements are made to the strategy itself.The main research contents are as follows.(1)For the problem of multi-fault diagnosis,two new algorithms are implemented by improving and combining two-population intelligence algorithm,information entropy and Huffman coding.Compared with the traditional single test sequence,in order to meet the complex requirements of multi-fault diagnosis,another auxiliary sequence is used to precisely lock the diagnosis path of each multi-fault source in the optimization process.The experimental case shows that this method can ensure the fault isolation rate and reduce the test cost accurately.(2)In order to improve the practicability of multi-fault diagnosis strategy,the function of adjusting the number of source faults is realized.For example,the branch on the right corresponds to the failure of all units at the same time,but it is almost impossible to occur in reality.Therefore,this paper proposes a diagnosis method that can adjust the maximum number of source faults to avoid the exploration on unnecessary branches.The results show that the algorithm not only performs well in multi-fault diagnosis,but also performs effectively in single fault diagnosis when the number of source faults is reduced to 1.(3)Developed testability analysis software.Through the software,the process from modeling to diagnosis strategy is more complete and has a friendly human-computer interaction interface.The feasibility of the algorithm and the software is proved by the case verification.