Research On Optimization Method Of Testability Design For Electronic System

With the increased integration of electronic products,electronic system grows more and more complex.This will inevitably leads to a higher and higher demand for the testability requirements of electronic system.The research of electronic system testability is mainly manifested in two aspects,namely,the allocation of the whole system test resources,and the improvement of the efficiency and accuracy of the research methods.It is usually in the distribution of the all test resource and the improvement of accuracy and efficiency of testability research methods,which plays a crucial role in insuring the fault diagnosis ability of electronic system and reducing the cost of electronic system test and maintenance.In order to improve the detectability of the system,it is necessary to incorporate testability into consideration at the beginning of the design of the system.The research conducted in this paper is listed as follows:1.The testability modeling method of electronic system is briefly introduced,which is the basis of testability study.This paper focuses on the multi signal model,including the concept,method and advantage of the multi signal model,and an example is given to introduce how the multi signal model models the electronic system.In addition,the testability model of electronic system is also introduced.2.This paper gives a brief description of the sequential test problem and its commonly used optimization algorithms,and a sequential test optimization algorithm based on improved genetic programming is proposed.Sequential testing is a classic problem of the research on testability of electronic system at home and abroad,and almost all of the solutions to it are based on heuristic search algorithms.The AND/ OR graph search AO* algorithm is taken as an example in this paper,for introducing the heuristic search algorithm.The disadvantage of heuristic search algorithm is that no matter how to improve the algorithm,it is necessary to construct complex heuristic function.Besides,the computational complexity increases exponentially with the number of available tests of electronic system.In this paper,the idea of evolutionary algorithm is used to solve the problem of sequential test,and a sequential test optimization algorithm based on improved genetic programming is proposed.The basic ideaof this algorithm is to develop an optimal or near optimal sequential fault search strategy from a group of randomly generated fault diagnosis tree.The optimization process is simple and easy to understand.It breaks the limitation of the traditional heuristic search algorithm,and the experimental results show its optimization effect is satisfactory.3.This paper gives a detail description for the problem of electronic system testability under the condition of unreliable test,and puts forward a method of packet coding based genetic algorithm to solve it.First of all,the practical significance of this problem is introduced,and then several traditional ideas to solve the problem are described.The host intelligence algorithms used in this paper are genetic algorithm.Finally,two ideas of test optimization are put forward in this paper,and they are genetic algorithms based on decomposition-parallel and block coding,individually.In addition,this method is also compared with the discrete particle swarm optimization(PSO)algorithm based on block coding to prove the validity of the proposed test optimization idea.