Research On Test Case Automated Generation Based On Artificial Immune Algorithm

With the development of software scales, the possibility of making errors is increasing. As a result, people attach great importance to the significance of enhancing the quality and reliability of software. And software testing is a necessary means of ensuring its quality and reliability. The method of test case generation is the core content of software testing research. Since the manually designed test case not only is energy-consuming and time-consuming but also is liable to errors, which leads to the high cost of software testing .Therefore, the research on test case automated generation becomes worthwhile which will be wildly used to reduce testers' workload and hence, the cost of tests.At present, there are various technologies of test case automated generation, among which the research focus is the application of computational intelligence technology for effective test case generation which includes the method of Artificial Immune Algorithm. It has good optimization capability and has been successfully used in many optimization calculation problems. This research aims to apply Artificial Immune Algorithm to handle test case automated generation problems. In the mean time, it also aims to find multiple test cases for multiple paths by using the optimization capability of Artificial Immune Algorithm. In our research, the following aspects have been researched:1. The research on the basic technology of test case generation and Artificial Immune AlgorithmFirstly, it introduces the approach of function-oriented test case generation and the approach of structure-oriented test case generation, especially, it focus on the research on the approach of path-oriented test case generation of the latter approach. Secondly, it also introduces the biology immune system's mechanism, features, functions and the basic framework of Artificial Immune Algorithm. Lastly, it gives a brief introduction of program instrumentation theory.2. The research on test case automated generation based on Artificial Immune AlgorithmOn the foundation of research on the relationship between test case generation and the immune system's conceptions and principles, this thesis proposes one approach of test case automated generation based on Artificial Immune Algorithm. It not only introduces the basic steps of testing case generator system but also analyses the parameters encoding and makes design schemes of affinity operator, concentration operator and stimulation operator, etc. Further more, the paper depicts the approach of test case automated generation based on Artificial Immune Algorithm.3. the experiment results and the analysisAccording to the Artificial Immune Algorithm, an experimental tool for automatically generating test cases is developed. Triangle classifier is chosen as the tested program. The author made experiments and made analysis of experiment results after program static analysis and program instrumentation.