论文部分内容阅读
混合整数非线性规划问题存在于大量工程和管理中,针对此问题提出一种滤子混合协同进化算法.利用滤子技术代替罚函数处理约束条件,采用混合编码和由差分进化算法与遗传算法异构的种群协同解决混合整数变量问题,引入基于平均熵和Logistic混沌初始化增加算法鲁棒性,利用自适应缩放因子和精英交流学习策略构成策略协同,与种群协同耦合,以提高算法搜索能力.以IEEE30节点测试系统进行无功优化为例,仿真结果表明所提出的算法具有全局搜索能力和有效性.
Hybrid integer nonlinear programming problems exist in a large number of engineering and management, and a filter hybrid co-evolutionary algorithm is proposed to solve this problem. The filter technique is used to deal with the constraint instead of the penalty function, the hybrid coding and the differential evolution algorithm and genetic algorithm The population of symbiosis solves the problem of mixed integer variables, and introduces the robustness of algorithm based on average entropy and Logistic chaos initialization, and uses the adaptive scaling factor and elite communication learning strategy to form the strategy synergy and synergistic coupling with the population to improve the algorithm search ability. IEEE30 node test system for reactive power optimization as an example, the simulation results show that the proposed algorithm has global search capability and effectiveness.