基于多智能体遗传算法的配电网节能降耗综合管理系统∗

摘要

针对目前企业配电网节能技术的不足,提出了一种基于多智能体遗传算法的配电网节能降耗综合管理系统。结合遗传算法(Genetic algorithm,GA)和多智能体系统(Multi-Agentsystem,MAS)技术构造了一种 GA-MAS 算法,每一个多智能体相当于遗传算法中一个个体,相邻的多智能体相互作用,并结合遗传算法的进化机理进行全局最优求解。提出了该系统各节能设备智能体结构模型和高压/低压多智能体系结构模型,运用 GA-MA S算法,得出各个节能设备的最佳调节力度,使节能设备以最小的调节代价获得最大的节能效益。具体算例仿真及工程实际应用表明本文提出的配电网节能降耗综合管理系统能使总有功网损降低,电容器投入总组数减少,实现节能设备的最佳调节,同时表明 GA-MAS算法收敛速度较快。%Aiming at the deficiency of energy saving technology in enterprise distribution network,an energy saving management system of distribution network based on multi-agent genetic algorithm was pro-posed.Combined with genetic algorithm and multi-agent system,a GA-MAS algorithm was proposed. Each multi-agent was equivalent to an individual of genetic algorithm,and the adj acent multi-agent was in interaction.The GA-MAS algorithm was combined with the evolutionary mechanism of the genetic algo-rithm for global optimal solution.The agent structure model of the energy saving equipment and the intel-ligent architecture model of high/low voltage system were presented.Using the proposed GA-MAS algo-rithm,the optimal regulation of energy-saving equipment was obtained,so the least cost of the energy-sav-ing equipment had the biggest energy saving profit.The simulation and practical application have shown that the proposed energy saving management system of distribution network can reduce the total activepower loss and the total number of capacitors,and achieve the best regulation of energy-saving device.It has also been shown that the proposed GA-MAS algorithm has faster convergence speed.