The Study Of Dynamic Optimization And Multi-Objective Optimization Algorithm In Sequential Test Problem

As the electronic systems become increasingly complicated,it is more difficult to test their states and performances.Therefore,testability needs to be considered and included into the design system.The purpose of the sequential test problem is to construct the test sets with minimum test costs to identify system fault,which reduces maintenance costs of the systems.This paper makes a deep research on sequential test problem in different aspects.Testability design and analysis software is short in this field,and most of them are based on client /server architecture.Our software implement relevant functions based on browser/server architecture.The main work of this paper is as follows:1.This paper introduces the sequential test problem and the most used algorithm for this problem: AO* algorithm.Since the performance of the AO* algorithm mainly depends on the choose of the heuristic evaluation function,two different heuristic evaluation function is discussed.The heuristic evaluation function based on Huffman coding can obtain the optimal test costs of the systems.The heuristic evaluation function based on entropy cannot obtain the optimal test costs but the efficiency is much higher than that of the heuristic evaluation function based on Huffman coding.The differences and the use scope of these two heuristic evaluation function based on two experiment example are also provided.2.The parameters of the sequential test may change throughout the life cycle of the systems.Under these circumstances,instead of rerunning the whole AO* algorithm thoroughly,we make trivial adjustments on previous decision tree to accommodate the new circumstance.This method is much more efficient than the traditional AO* algorithm.Besides,the decision tree can evolve with varying environment and maintain history.Without loss of accuracy,the time efficiency is improved.3.The multi-objective optimization problem and combine of the sequential test problem and the multi-objective optimization problem are introduced at first.To solve this problem,the genetic programming algorithm based on multi-objective peaks is proposed.This algorithm has the classic genetic programming operations and some operations for multi-objective optimization as well.The output of this algorithm is a group of non-dominated solutions for system reference.4.This paper introduces the overall design process of the software and the technology and data structure of the software development course.Then the major function modules are introduced in detail.The system modeling module can be used for automatic modeling or manual input the system parameters.The sequential test module can generate the decision tree and report of the system.The dynamic changing module provides modification of the original decision tree based on the change.The multi-objective optimization module can produce a set of non-dominated solutions.