Test Cases Generation Method Of High-order Usage Model Based On Determined Order And Test Adequacy Criteria

In recent years,various application software have emerged and the scale of software development has become larger and larger.As an important means to ensure high quality and high reliability of software before it is officially put into use,preparation for software testing becomes more and more important.The automatic generation of test cases becomes the key and difficulty of reducing test cost and improving test efficiency in the software test phase.The efficient automatic test case generation technology aims to generate as few test cases as possible to cover the software testing requirements as much as possible.Therefore,the difficulty of test case generation technology lies in how to generate high coverage test case data automatically,and how to determine an effective test stop criteria to end the test as soon as possible under the premise of ensuring the test is sufficient,thereby reducing test costs and improving test efficiency.Currently,there are many studies on the automatic generation of test cases and test adequacy criteria at home and abroad.Although some satisfactory results have been achieved,there are still some problems.For example,the test case generation method is unstable with low coverage and low efficiency,and the test adequacy criteria is inaccurate with poor stability and is prone to "premature" phenomenon.In order to solve the above problems,based on the High-order Markov usage model,this paper conducts research on the test case automatic generation and the adequacy test criteria.The specific contents include:(1)An automatic test case generation method based on Higher-order Markov usage model of the determined order is proposed.Compared with the test case generation method based on the First-order Markov usage model,the improved High-order Markov usage model solves its first-order inefficiency problem.The test case generation algorithm is more stable and has higher coverage and faster speed.(2)A test adequacy criteria based on weighted symmetric relative entropy is proposed.Compared with the test adequacy criteria based on the Discriminant value commonly used by Markov usage model,the improved test adequacy criteria can accurately measure the difference between Markov usage model use chain and test chain,and there will be no"premature" phenomenon,making the adequacy judgment more stable and accurate.(3)Based on the above test case automatic generation method and the test adequacy criteria,a general software automation test system was developed.Take a UAV ground control station system as an example to verify the feasibility and effectiveness of the overall algorithm.Experiments show that the method is practical and effective,and can be used as a good pre-selection scheme and research direction for software testing.