Optimization Model Of Path Selection For Software Testing And Its Evolution-based Solution

Software testing is an important means of software quality assurance,which runs through the entire software development life cycle.As a widely used method,path testing has the premise of testing effective paths.However,existing path generation methods have disadvantages such as low testing efficiency and easy generation of redundant paths,which makes it difficult for path testing to meet actual requirements.In addition,parallel programs are now widely used in large-scale scientific or engineering calculations,but the existence of multiple processes and communication statements in parallel programs further increases the difficulty of generating test path.In view of this,this article mainly studies the path selection optimization model used for software testing and its evolutionary solution method.The main research work of this paper is as follows:(1)Aiming at the problem that the existing path generation methods are easy to generate redundant test paths,a path selection optimization model for serial program testing and its evolutionary solution are proposed.To solve this problem,this paper establishes an optimization model and uses a multi-objective evolutionary algorithm to solve it.The main idea is: First,take multiple paths as decision variables,and establish a multi-objective optimization model based on the number of edges,paths and coverage difficulty contained in the decision variables;then,the multi-objective evolutionary algorithm is used to solve the model to obtain the set of goal paths that meet the needs.The proposed method is applied to 7 benchmark test procedures and compared with other 5 classical algorithms.Experimental results show that compared with other algorithms,the proposed method can reduce the number of redundant paths and reduce the difficulty of coverage under the condition of ensuring test adequacy,thereby reducing the time consumption of generating test data and improving test efficiency.(2)Aiming at the problem of increased testing difficulty caused by the existence of multiple processes and communication statements in parallel programs,a path selection optimization model and its evolutionary solution for parallel program testing are proposed.This method needs to establish a multi-objective optimization model suitable for parallel programs,and use multiple parallel program paths as decision variables,and determine the following four indicators: the total number of control edges and communication edges included in the decision variables,the number of parallel program paths,the proportion of the number of communication paths in the path set,and the coverage difficulty of the path set.Design two objective functions based on the above four indicators.In the process of solving the multi-objective optimization model,the population is evolved through the cross operation between and within the set until the set of target paths that meet the needs is obtained.The proposed method is applied to 7 message passing parallel programs and compared with other 3 classic algorithms.Experimental results show that compared with other algorithms,the proposed method can effectively select the test path of parallel programs.The above-mentioned research work has enriched the path selection theory and methods of serial and parallel program path testing,expanded the application range of multi-objective evolutionary algorithms,and has important theoretical significance and practical value.This article contains 10 figures,18 tables and 92 references.