摘要
介绍一种类似于遗传算法的进化算法———粒子群优化算法, 并把它应用到电力系统无功优化问题中。对基本的粒子群优化算法作了适当改进, 在粒子速度更新公式中增加了一项即上一代的全局“最优”值, 相当于增加了全局极值的权重, 提高了算法的收敛性。以粒子群优化算法为基础, 选取适合于该算法的无功优化目标函数。通过对 IEEE- 14节点的仿真计算, 证明了该算法优于基本的粒子群优化算法, 且与遗传算法相比能在更少的迭代次数内搜索到更好的全局最优解。
The Particle Swarm Optimization algorithm (PSO) is introduced and its application to reactive power optimization in power system has been investigated in the paper. To improve the search efficiency of PSO, a third extremum is insert to the iteration formula to indicate the search direction and a “fly back” strategy is proposed. Simulation results on IEEE-14-bus power system show that PSO can find better solution compared with GA and search ability of the modified PSO is better than standard PSO.
出处
《电力勘测设计》
2005年第1期66-70,共5页
Electric Power Survey & Design
关键词
粒子群算法
无功优化
收敛
全局最优
reactive power optimization
particle swarm optimization (PSO)
evolutionary computation
