小生境遗传算法在无功优化中的应用研究

在基本遗传算法（SGA）的基础上,引入生物学中小生境的概念,制定了初始种群生成方法,以保证个体的多样性,从而形成可用于电力系统无功优化的小生境遗传算法.应用此算法时,可用共享度改变个体的适应值,同时加速淘汰适应值低的个体,提高每一代个体的平均适应值水平,以减少迭代的次数.运用IEEE6节点系统和168节点实际电网进行计算的结果表明：在优化条件相同时,该算法的迭代次数明显少于基本遗传算法,提高了无功优化的收敛速度.

RESEARCH ON NICHE GENETIC ALGORITHM FOR REACTIVE POWER OPTIMIZATION

Based on SGA, this article introduces Niche that is a concept of biology, specifies the generative approach of initial population to guarantee individual diversity, and then forms the Niche Genetic Algorithm that can be used in reactive power system optimization. This algorithm changes individual adaptive value with sharing degree, accelerates to eliminate individuals which have low adaptive value, increases the average adaptive value of every generation, and achieves the goal of reducing iterative times. The proposed method is applied to IEEE6 buses and 168 nodes practal power system. The result of calculation shows that compared with SGA in the same condition, this approach can reduce remarkably iterative times and improve rate of convergence of the algorithm for reactive power optimization.