Improved algorithm for combinatorial test data generation based on particle swarm optimization

In this paper, an algorithm for test data generation, which is an important problem of combinatorial testing, based on particle swarm optimization (PSO) method was introduced. The decaying threshold mechanism was used to select all particles in the decaying range after every optimizing process. After every PSO iteration, the best particle was outputted as a test data and all these selected particles were disturbed by a new disturbance method which was based on the best data output. The results indicate that more test data can be outputted after every PSO calculation. Compared with the original algorithm, the times of optimization calculation was reduced effectively, and the efficiency of the test data generation was improved effectively on the premise of ensuring the optimization degree of the data generated.