Application Of Improved Simulated Annealing Genetic Algorithm In Software Test Case Generation

In the whole software development process,development and testing of software are important links,and are equally important.During development,testers are generally responsible for writing test cases,so the coverage of test cases on development content is not comprehensive enough,and the efficiency of testing is relatively low.Therefore,how to improve the generation efficiency of test cases has become an unavoidable problem for designers.To solve the above problems,this paper has put forward an improved Simulated Annealing Genetic Algorithm(SAGA),in which a new path similarity calculation is first proposed,and is put into the calculation of fitness function,the evaluation of individuals is more accurate;Secondly,multi-point crossover algorithm is selected around crossover operator,and adaptive mutation is adopted for mutation operator,and adds elite retention strategy in selection operation to avoid the destruction of the optimal individual due to other operations in the evolution;Finally,the improved algorithm is fully combined with the simulated annealing algorithm,and the Bolztmann mechanism is applied in the new algorithm to avoid premature and local optimal solutions,so that the convergence speed is faster and optimal.By using four benchmark programs and one practical program,this paper has compared and analyzed the research results of other relevant scholars from three aspects: running time,optimal iteration times and optimal solution generation,and has verified the reliability and effectiveness of SAGA in generating test case efficiency.Then,by taking the test environment of a software company as the actual background,a test case generation system has been developed.SAGA algorithm has been used in the system,and good application results were obtained,demonstrating the superiority of the new algorithm.