基于改进离散粒子群算法的测试优化选择

为了解决复杂系统测试性设计过程中测试选择所产生的组合爆炸问题，提出一种改进离散粒子群算法的智能方法。首先，为保证初始种群的多样性，利用混沌不重复遍历的特性初始化种群的速度和位置；其次，根据启发式规则和罚函数的方法计算粒子适应度，使算法具有良好的搜索性能；最后，通过采用自适应调整策略的惯性权重，使粒子易于跳出局部最优解，找到最优解。通过仿真实例验证了本文方法的有效性，优化结果满足系统各项测试性指标要求，可为复杂系统的测试优化选择提供有效指导。

Optimal Test Selection based on Improved Discrete PSO Algorithm

Due to the fact that the combinatorial explosion problem of test selection in the design for testability for complex system, an intelligent method of test selection based on an improved discrete particle swarm optimization algorithm (IDPSO) is put forward. First, the ergodic of chaos has been used to initialize the velocities and positions of the particles. Then, the fitness value is calculated by heuristic rule and penalty function to improve the search performance. Finally, The inertia weights were adjusted according to adaptively strategy, which can avoid the particles trapped in local optimal. The simulation results meet all system requirements and show that IDPSO algorithm can achieve global optimal solution fast and effectively, which makes it a good solution to the optimal test selection for complex system.