Research And Implementation Of Automatic Test Cases Generation Of Airborne UAV Flight Control Software

The research of UAV is in the ascendant, and its position in national defense and civil defense is more and more prominent, which makes the development of UAV more and more attention. As one of the most important part of the UAV flight control system, it will affect the safety of UAVs once the failure.The quality of the control system software directly affects the safety of flight control system. Therefore, in the design of flight control system software, how to carry out effective comprehensive testing software is indispensable. It has important practical significance to study the automatic generation algorithm of test cases, and to realize the automatic generation of test cases by software engineering.In this paper, the theory of automatic test case generation is studied with the UAV flight control system as the research object, combining with SCADE software being used in a Chengdu research institute. The advantages and disadvantages of several algorithms are compared and analyzed. The idea of fusion of two algorithms is presented to avoid the disadvantages of each algorithm. It can make sure that obtaining a higher coverage rate in case of fewer test cases automatic generated. Finally, engineering implementation of the proposed scheme, and testing the coverage effect of test case set.Firstly, the generation algorithm of test case automatic generation is deeply studied, including GA, Ant Colony Algorithm and combined test generation algorithm. The differences and application scope between these algorithms are introduced, and stressing on principle and defect of the combined test algorithm. Then it introduces the IPO generation algorithm, the concept of causal graph constraint algorithm. According to the characteristics of the combination test algorithm and the causality constraint algorithm, the combination of the two algorithms is carried out. Then theory of automatic generation of test cases in this paper is proposed.Secondly, this paper introduces the specific implementation scheme of the automatic generation of test cases based on the theory, including randomized design, counterexample generation algorithm design and so on. Some more targeted treatment designs are implemented in design process according to the characteristics of SCADE modeling software and flight control system program. Finally, a test case automatic generation tool based on SCADE modeling software platform is developed.Then, the test case generation tool is tested and verified. The different models are compared and validated from different parties, mainly including the ideal model and the flight control subsystem models. Manual and automation are compared and the coverage ratios between one generating algorithm and two algorithms are compared as well. At last, test and statistical results are listed.In conclusion, through the theory to the project implementation, and ultimately to the verification process, this paper realizes the test case automation generation. It achieved an satisfactory result on the aspect of reducing the artificial workload in the test case design process.