Test Data Generation Using Annealing Immune Genetic Algorithm

With the development of software technology and the expansion of software project scale, software testing appears to be more crucial. And test data selection is one of the nodi during software structure testing because the suitability of test data may directly affect error detection. Notwithstanding existence of several methods to generate test data automatically, such an algorithm overcoming disadvantages of the existing methods in practice hasn´t been brought out, that some errors still have to be detected by engineering experience. Therefore, this paper analyzes the characteristics and shortcomings of simple genetic algorithm, simulated annealing genetic algorithm as well as immune algorithm respectively. Aiming at solving the shortcomings in standard Genetic Algorithm on search efficiency, individual diversity and premature, the Annealing Immune Genetic Algorithm (AIGA) is presented as the core algorithm of test data generation by introducing the mechanism of reproduction rate adjustment of individual concentration of immune algorithm and annealing principium into genetic algorithm. Finally, AIGA mentioned above was applied and verified with a practical software testing example.