Research On Test Case Generation Of Multi-UAV Task Distribution System

With the rapid development of the Internet,the demand for software is increasing,and the design of the system is gradually complicated.Therefore the difficulty of combinatorial testing are growing up.The purpose of the combined test is to test the system with as few test cases as possible to detect possible problems with the system.One of the research hotspots of combinatorial testing is the generation of combinatorial test suite which is to generates as few test case sets as possible that satisfies the given portfolio coverage requirements.Before combinatorial testing,it is necessary to extract the constraint relationship in the system which to be tested.If the constraint relationship in the system software is not fully extracted,some invalid test cases may be generated to reduce the efficiency of the software test work.Because of the many constraints that need to be considered in the multi-UAV task distribution system,the constraint relationship extraction is an important issue before testing the UAV task allocation system.Based on the above problems,the main work of this paper are as follows:(1)In this paper,we propose a method of extracting constraint relations based on semantic reasoning.In the multi-UAV task assignment system,constraints between UAV and tasks may become different at different moments.Therefore the constraint relationship of factors in the system is summarized by semantic reasoning,based on the actual value,the combination set that need to be covered is generated.The size of the coverage combination set is reducing and avoid invalid combinations;(2)In this paper,we propose a new priority measurement function based on combined coverage.The function prioritize the combination of the combination set which needs to be covered and is generated in the first step,and select the combination with the highest priority to make the selected combination coverage higher.(3)In this paper,we propose a test case generation method based on improved particle swarm optimization.The method combines the improved particle swarm algorithm with the one-test-at-one-time strategy to determine the pending values in the selected combination of the previous step,and generates a test case set for the multi-UAV task assignment system.The improved particle swarm optimization algorithm combines the affinity theory of artificial immune algorithm,and uses the information entropy calculation method to evaluate the aggregation degree of particles.This enhances the "premature" problem of the particle swarm algorithm which is caused by the reduction of particle diversity in the middle and early iterations,and reduces the number of test cases generated.(4)In this paper,we build an online test to perform online testing of the multi-UAV task distribution system.In the platform,we use the selenium automated test framework to convert the test case set which are generated in the third into a test script,and test the multi-UAV task distribution system automatically.The experimental results show that the proposed method based on semantic reasoning can successfully infer the implicit relationship between the UAV and tasks in the multi-UAV task assignment system,and the constraint conditions based on the constraint relationship become comprehensive.The proposed priority measurement function prioritize the combinations of the Combination set that needs to be covered.The function avoids random selection and enhances the quality of the selected combination.The improved particle swarm algorithm selects an appropriate value for the selected combination which has the value to be determined,and can get a small number of test case sets with a wide coverage.