摘要
简要分析了几种无功优化方法的局限性,通过比较得出遗传算法是求解无功优化的一种有效的方法,并在简单遗传算法(SGA)的基础上,提出了更加有效的算法即改进遗传算法(IGA)。该算法针对常规遗传算法收敛速度慢、易早熟等缺陷,并结合电力系统无功优化的特点,借鉴了模拟退火思想在遗传算法的执行过程中对个体适应度不断进行修正,并采用了浮点数编码、双层结构群体、自适应的交叉率和变异率等改进措施。算例表明这种改进的遗传算法优化效果好,而且在精度和收敛度上都有较大提高。
This paper briefly analyzes the limitation of some reactive power optimization methods for power system and through comparison draws the conclusion that genetic algorithm is an effective method for reactive power optimization. A more effective method ——the improved genetic algorithm is put forward based on simple genetic algorithm. The new algorithm conquers the defects of convergence rate slow and easily being precocious of simple genetic algorithm, and links up the characters of reactive power optimization of power system, and profits from the simulated annealing in the execution process of genetic algorithm to amend the individual fitness constantly. It applies float point encoding, muti-population, self-adaptive variation rate and self-adaptive crossing rate as well. The calculation example indicates that this improved genetic algorithm has the optimal results with improved precision and convergence speed.
出处
《继电器》
CSCD
北大核心
2006年第13期29-32,53,共5页
Relay
关键词
无功优化
改进遗传算法
自适应
浮点编码
双层结构群体
reactive power optimization
improved genetic algorithm
self-adaptive
float point encoding
multi-population
