基于混沌免疫接种粒子群算法的电力系统多目标无功优化

Multi-objective reactive power optimization based on the algorithm of chaotic immune-vaccine particle swarm in electric power system

针对基本粒子群优化算法的收敛性能受初始粒子分布影响较大的问题,应用混沌优化理论具有遍历性的突出优点和人工免疫系统中接种疫苗的方法,即混沌免疫接种粒子群优化算法,将该方法应用于电力系统多目标无功优化模型求解。混沌免疫接种粒子群算法是采用混沌优化生成初始粒子即无功优化控制变量值,并选择其中较优的粒子作为初始粒子群,改善了随机初始化过程无法保证粒子合理分布的缺点;同时,采用人工疫苗接种优化技术提高其搜索速度和精度。通过IEEE 30节点系统算例仿真研究,并分析比较基本粒子群算法和改进后的粒子群优化算法的无功优化模型求解。结果表明,应用混沌免疫接种粒子群算法来解决多目标无功优化问题是有效可行的。

To reduce the influence of initial particle distribution on convergence of basic particle swarm optimization algorithm,this paper proposes to solve multi-objective reactive power optimization model in power system by applying chaotic immune-vaccine particle swarm optimization algorithm which enjoys the advantage of ergodicity of chaotic optimization theory and immune vaccine of artificial immune system.The algorithm guarantees the rational particle distribution in random initialization by adopting chaotic optimization to generate initial particle,which is the control variable of reactive power optimization and choosing the optimal particles as initial particle swarm;at the mean time,the algorithm adopts artificial immune-vaccine optimization to enhance the rate and accuracy of research.Through the simulation study on IEEE30 joint system example,this paper analyzes and compares reactive power optimization model solution of both basic and improved particle swarm optimization algorithm.Analysis proves the feasibility of chaotic immune-vaccine particle swarm optimization algorithm for solving multi-objective reactive power optimization.