Research On Test Case Related Problems Based On Swarm Intelligence Algorithm

With the rapid development of IT and society informationization,the software industry has become more and more mature,and there are more and more large-scale software systems being built.At the same time,the requirements for software quality have become higher.Software testing is an indispensable link in software quality and technology.It has an irreplaceable position in the entire software development cycle.Regression testing is a common and important method in software testing.It is mainly applied quality assurance after software changes.However,with the increase of software scale,the cost of regression testing also increases,and its proportion in the software testing process is increasing.Therefore,in order to reduce the test cost and improve the utilization of resources under the condition of ensuring quality,it has become a very important research direction for regression testing.An important way to improve the efficiency of regression testing is to optimize the test suite.There are currently three common optimization methods,which are test suite reduction,test case selection,and test case prioritization.These methods are applicable to different scenarios.This article mainly focuses on test suite reduction and test case prioritization.After understanding the deficiencies of traditional methods,this article introduces the widely used swarm intelligence algorithm to the test suite optimization problem,Proposed a new test case reduction and test case prioritization technology:The first algorithm is the test suite reduction technique based on the good-point-set firefly algorithm.The test suite reduction problem belong to the NP-Hard problem,so the traditional greedy algorithm is difficult to find the optimal solution,and the integer programming method with strong solvency is too time complex to be applied in actual production.To solve these problems,this paper introduces the firefly algorithm with fast convergence speed and strong traversal ability.In order to solve the problem of poor search ability of the algorithm,the good point set is uesd for optimization.Afterwards,binary coding was applied to the test suite reduction problem.By analyzing the experimental results,the firefly algorithm with the good point set has been improved in terms of solving ability,and this paper proposes that the good-point-set firefly algorithm can make a better test suite reduction set.The other algorithm is a test case prioritization technique based on a multi-target bat immune algorithm.Test case priority ranking problem often has multiple indicators,so it is a multi-objective optimization problem.This paper introduces the Pareto optimal solution set to test point coverage rate and maximum coverage cost.To solve this type of problem,the bat algorithm in swarm intelligence algorithm is applied.The algorithm has strong spatial search ability,but it may have problems of premature convergence.In order to solve these problem,the artificial immune system is used to optimize the algorithm.,which can improve the diversity of bat particles.Finally,the traditional coding method cannot be directly applied to the test case set ranking problem,so this paper uses a genetic coding method.After the experiments,the algorithm can obtain a better the Pareto optimal solution with the test case priority problem.