摘要
针对免疫算法在配电网重构中收敛速度慢,易收敛到局部最优解等问题,提出矢量距浓度免疫算法。抗体的亲和度决定它的矢量距,并由矢量距得出抗体的浓度、选择概率和期望繁殖率,根据选择概率自适应调整抗体的变异率,根据期望繁殖率进行克隆操作以保证抗体的多样性和全局最优解的生成,最后结合疫苗接种和免疫记忆机制促进全局最优解的生成。算例结果表明,该算法能有效提高收敛速度和保证全局最优解的生成。
A new immune algorithm based on vector distance density is proposed to solve the problems that immune algorithm often faces in network reconfiguration,such as slow convergence and easy convergence to local optimal solution.The vector distance is decided by the antibody's affinity,and the antibody density,select probability and expected reproductive rate of antibody obtained.The antibody mutation rate was adaptive according to the selection probability,and the diversity of antibody and global optimal solution is ensured by the cloning operation which was based on expected reproductive rate.Finally,the combination of vaccination and immune memory mechanisms can promote the formation of the global optimal solution.The example verifies the algorithm that can enhance convergence speed and ensure the global optimal solution' generation.
出处
《电力系统及其自动化学报》
CSCD
北大核心
2012年第1期79-83,共5页
Proceedings of the CSU-EPSA
基金
国家科技支撑计划项目(2008BAA13B01)
关键词
矢量距浓度免疫算法
配电网重构
变异
克隆
疫苗接种
免疫记忆
vector distance based immune algorithm(VDIA)
distribution network reconfiguration
mutation
cloning
vaccination
immune memory