Test Case Generation And Prioritization Based On Artificial Bee Colony Optimization Algorithm

With the development of various software products in our daily life,the people have higher demands for software quality.The quality of software is ensured by kinds of software testing technologies.Among them,the automated test case generation technology and test case prioritization technology can improve the testing efficiency of software and reduce the testing cost.The existing research shows that the meta-heuristic search algorithm can effectively solve the problem of automatic generation and prioritization of test cases.But the research mainly focuses on a few kinds of algorithms.There is still room for improvement among them.Meanwhile,there are few studies on new search algorithms.Therefore,the research on the automatic generation and prioritization of test cases still has theoretical significance and practical application value.This paper applies the artificial bee colony optimization algorithm to test cases automatic generation and prioritization.The main research contents are summarized as follows:(1)Aiming at the artificial bee colony(ABC)algorithm exist the problem of searching slow in the early stage,this paper combines the genetic algorithm(GA)with it.Because of the GA has a faster searching speed in the early stage.An improved adaptive genetic-artificial bee colony(IAG-ABC)algorithm is proposed in this paper.The IAG-ABC algorithm is applied to generate testing cases for path coverage.(2)Aiming at the multi-objective artificial bee colony(MOABC)algorithm exist the problem of falling into the local optimal solution,this paper proposed a multi-objective artificial bee colony optimization(MOABCO)algorithm.At first,set the external elite solution set and the global optimal update strategy.Then,set the optimal solution-guided differential variation local search strategy.Finally,set the honey source selection strategy based on information entropy.Which improve the MOABC algorithm's searching performance.(3)This paper combines the average sentence coverage,effective execution time and historical defect discovery rate as the optimization objectives of the test case prioritization.However,the calculation complexity of three objectives optimization problem is high.But the effective execution time and historical defect discovery rate can be combined.The concept of historical defect detection efficiency is proposed in this paper.The three target optimization problems are transformed into two optimization targets.Finally,apply the MOABCO algorithm to solve the test case prioritization problem.The experimental results show that the IAG-ABC algorithm has higher convergence speed and global optimization performance than the existing adaptive genetic algorithm and bee colony algorithm in solving the problem of automatic generation of test cases.The MOABCO algorithm has higher convergence speed and higher defect detection rate than NSGA-II algorithm and OMOPSO algorithm when solving the test case prioritization problem in regression testing.