摘要
粒子群算法在电力系统无功优化中容易出现局部收敛和早熟的现象,对传统粒子群算法进行了改进。改进算法对当前全局最优粒子进行临近搜索(NSGOP-PSO),加快了粒子群的搜索速度和搜索精度。并加入了仿生进化算法中的灾变算法(NSGOP-CPSO),按一定规律对群体全部粒子或部分粒子施行灾变,用于解决粒子群算法中的不收敛和早熟现象的问题。对该算法的作用原理进行了细致的分析和说明,并以IEEE-30节点系统的仿真图形及统计数据说明了该改进粒子群算法是有效的。
In order to resolve the issues of easy prematurity and partial convergence of basic particle swarm optimization （BPSO） applied to reactive power optimization in power system, an improved BPSO is proposed. Neighborhood search of global optimal particle is added to PSO （NSGOP-PSO）, which improves the search speed and precision of particle swarm. Moreover, catastrophic algorithm based on evolutionary algorithm is also added （NSGOP-CPSO） in the whole particles or parts regularly, which can solve issues of prematurity and non-convergence of BPSO. Detailed analysis and explanation of action principle of this algorithm, simulation diagram and statistical data of IEEE30 nodes system show the effectiveness of NSGOP-CPSO.
出处
《电力系统保护与控制》
EI
CSCD
北大核心
2011年第18期110-115,共6页
Power System Protection and Control
关键词
无功功率优化
粒子群算法
临近搜索
灾变
作用原理
reactive power optimization
PSO
neighborhood search
catastrophe
theory of action
