基于改进离散差分进化算法的测试选择

Test Selection Based on Enhanced Discrete Differential Evolution Algorithm

摘要:

针对武器装备系统的测试选择问题,首次提出利用基于混合策略的离散差分进化算法(MD-DDE)进行测试选择；重新定义种群中个体的含义,根据其容易陷入早熟的特点,引入自适应变化的缩放因子和交叉因子；采用混合差分策略,进一步平衡了算法的全局寻优和收敛速度的矛盾；在选择操作的基础上突出了优先级的特点,使结果更具应用价值；通过实例和离散粒子群算法进行了对比,验证了该算法的有效性。

Abstract:

For test selection of weaponry system, a discrete differential evolution algorithm (MD-DDE) based on a hybrid strategy is proposed firstly. Redefine the meaning of the individuals of population, according to the characteristics of its tendency to fall into premature, adaptive changes in the scaling factor and crossing factor is introduced. By hybrid differential strategies, the contradictory of global optimization and convergence rate is further balanced. The feature of priorities is highlighted based on the selection operation that makes results more valuable. Through a example discrete particle swarm algorithm is compared to verify the effectiveness of the algorithm.