基于多Agent粒子群优化算法的电力系统经济负荷分配  

An economic load dispatch method of power system based on multi-Agent particle swarm optimization algorithm

作　　者：唐贤伦[1] 张衡[1] 李进[1] 周维[1] 

机构地区：[1]重庆邮电大学工业物联网与网络化控制教育部重点实验室,重庆400065

出　　处：《电力系统保护与控制》2012年第10期42-47,共6页Power System Protection and Control

基　　金：国家自然科学基金项目(60905066);重庆市自然科学基金项目(sstc2011jjA1313)~~

摘　　要：为更好地解决电力系统中经济负荷分配非连续、非凸、非线性的优化问题,提出一种新的基于多Agent系统竞争与合作机制的多Agent粒子群优化算法。该算法以粒子群算法和智能体技术为基础,将粒子群算法的群搜索特征和智能体的智能搜索特征相结合,使粒子在搜索过程中利用群体信息,环境信息共同决定搜索策略。每一个Agent相当于粒子群优化算法中的一个粒子,通过与其随机选取的邻居的竞争、合作操作,能够充分获得问题信息并自适应地权衡粒子的全局探索能力和局部开发能力,可更快、更精确地收敛到全局最优解。对IEEE3节点,IEEE13节点和IEEE40节点系统的仿真结果表明,该算法在求取电力系统ELD问题时具有较高质量的解和运行速度。An improved multi-Agent particle swarm optimization algorithm （RN-MAPSO） based on multi-Agent system competition and collaboration mechanism is proposed for solving the discontinuous, non-convex and nonlinear economic load dispatch（ELD） problems of power system. The algorithm combines the swarm search feature of the PSO with the intelligent search feature of the agent based on the particle swarm optimization algorithm and technology of agent. In the search process, each agent is treated as one particle of PSO that can take advantage of the group information and the environmental information to determine the search tactics. By competition and cooperation with the randomly selected neighbors, the agent can adaptively adjust its global searching ability and local exploring ability, and converge the global optimal solution more accurately at higher speed. To verify the effectiveness of the proposed algorithm, RN-MAPSO is tested in the IEEE 3 nodes, 13 nodes and 40 nodes system and the experiment results show that the proposed algorithm for ELD problems can acquire high-quality solutions rapidly.

关 键 词：电力系统 经济负荷分配 粒子群优化算法 多AGENT系统 竞争 合作 

分 类 号：TM714[电气工程—电力系统及自动化] TP18[自动化与计算机技术—控制理论与控制工程]