基于混合离散二进制粒子群-遗传算法的测试配置方法研究

Research on Test Configuration Method Based on Hybrid Discrete Binary Particle Swarm Optimization -Genetic Algorithm

摘要:

针对目前测试性建模工作中尚无具体方法指导测试配置这一问题，通过对系统内故障传播关系进行分析，提出了一种混合离散二进制粒子群-遗传算法用于求解测试配置的最优方案，使系统的测试性模型在满足规定测试性指标下使用的测试数量最少。将系统测试配置方案进行二进制粒子编码，并在粒子群算法中引入遗传算子，使混合算法具有较快的搜索速度的同时避免陷入局部最优。最后通过实例计算与仿真，证明所提出算法计算结果正确且对于指导复杂系统测试性建模工作具有实际应用价值。

Abstract:

In view of the present testability modeling work, there is no specific methods to guide the test configuration, the analysis of system failure propagation relation, this paper proposes a hybrid discrete binary particle swarm - genetic algorithm for solving the optimal scheme of test configuration, make the system testability model under the prescriptive testability index using the least number of tests. Binary particle coding is carried out in the system test configuration scheme, and genetic operators are introduced in particle swarm optimization algorithm, so that the hybrid algorithm has fast searching speed and avoids getting trapped in local optimum. In the end, it is proved that the proposed algorithm is correct and has practical application value to guide the test modeling of complex system.