粒子群优化算法精度和收敛性改进及应用

Study on Particle Swarm Algorithms for Smart Grid Dynamic Reactive Power Optimization

当前电网配电系统动态无功功率优化中,粒子群算法的精度和收敛性较低。提出一种基于粒子和惯性权重优化的改进粒子群算法。此算法首先调节粒子的状态,然后采用S曲线拟合来优化惯性权重在粒子群算法里的搜寻能力。最后,用此最优的学习策略来提高粒子群优化算法的收敛性。通过仿真验证,提出的基于粒子状态和惯性权重优化的粒子群改进算法比传统的粒子群优化算法具备更好的收敛性。此算法在电网配电系统中运行中产生的误差较小。

Due to the Iow accuracy and convergence of existing particIe swarm aIgorithm in the power grid dynamic reactive power optimization in distribution system,this paper proposes an improved particIe swarm aIgorithm based on the state of the particIe and inertia weight optimization.This aIgorithm first adjusts the states of the particIes.Then using Sigmoid map-ping to optimize the search abiIity of the inertia weight in particIe swarms aIgorithm.FinaI y,using the optimaI Iearning strate-gies to improve the convergence of particIe swarm optimization aIgorithm.