Research On Test Case Generation And Sorting Based On Meta-Heuristic Algorithm

Software test case generation technology and prioritization technology are two key technologies of software test automation.Meta-heuristic search algorithm is widely used to solve the problem of automatic generation and prioritization of test cases.This paper systematically studies and summarizes the existing research results in related technologies at home and abroad.It is found that the application of metaheuristic search algorithm in test case generation technology and prioritization technology is not mature,and the ubiquitous problem is that the algorithm converges slowly,considers the influencing factors to be single,and is difficult to converge to the global optimal.Excellent and consider the single factor of influence.To this end,this paper mainly studies the problem of meta-heuristic search algorithm for solving test case generation and prioritization,and proposes a dynamic boot test case generation strategy based on genetic optimization algorithm,And a multi-objective prioritization method based on dynamic reduction of ant colony optimization algorithm.The main research contents and specific contributions of this paper are mainly as follows:(1)In the aspect of test case generation technology based on path coverage,this paper uses a widely used genetic algorithm to solve.Considering the coverage of the initial test data on the path nodes,we first distinguish the difficult coverage paths,then design a path similarity calculation formula,analyze the heuristic information between the difficult coverage paths and use it to replace the partial initial of the genetic algorithm population.(2)In the improvement of genetic algorithm,the influence of branch weight on population fitness is considered,and each of the impact factors is weighted according to the characteristics of different programs,and a fitness evaluation function with weight influence factor is constructed.Adapt to the genetic probability and direct the individual cross-variation to quickly obtain high-quality test data that meets the path coverage.(3)In the test case prioritization technology,this paper uses the robust ant colony algorithm to solve this problem,combined with a dynamic reduction idea in the sorting process,and initially reduces the test case according to the demand coverage.Then,considering the error level detected during the actual execution of the test case,the method of judging the failure degree of the test case is designed to perform a second reduction on the test case in the iterative process,The time required for the ant colony algorithm iteration is greatly reduced by two reductions.In the pheromone update strategy of ant colony algorithm,this paper comprehensively considers the influence of test factor importance,failure degree and effective execution time on pheromone,and guides the update of ant colony pheromone online to enhance ant colony.The algorithm's solution accuracy and convergence speed.The sequence of test cases that the ant colony walks through in turn is the final sort result.In order to verify the effectiveness of the improved method proposed in the two aspects of test case generation and prioritization,multiple benchmarks and industrial programs are selected for programming experiments.The test case generation and multi-target prioritization methods based on the meta heuristic search algorithm proposed in this paper are compared with other methods.The simulation experiment results show that the genetic algorithm-based test case generation method based on this topic has obvious advantages in convergence speed,path coverage rate and utilization of existing test data.The proposed multi-objective prioritization method based on ant colony optimization algorithm is superior to other methods in terms of statement coverage,defect detection efficiency and effective execution time.