摘要
比较了遗传算法中二进制编码和十进制编码,指出后者更适合于无功优化问题,并结合无功优化问题的特点对遗传算法进行了改进:①采用十进制整实数混合编码方式;②研究适合十进制遗传算法的变异方法,并借鉴梯度概念,提出了基于进化方向的定向变异遗传算法;③采用映射法计算适应度函数,该方法简单易行,不会出现负值,能给群体施加定常的选择压力;④对每一代的最优个体进行单独变异,使搜索效率提高。该方法在甘肃金昌地区电网的无功优化计算中,效果较好。
After the comparison of the merits and demerits between binary and decimal coding, it is pointed out that decimal genetic algorithm (GA) is more suitable to the optimization of such a large scale, multi-variants and mixing integer optimization problem as the reactive power optimisation. Here, considering the features of reactive power optimization, following improvements of GA are performed, they are: 1.A grouped, integer and real number mixed coding method is applied in GA; 2.The mutating method suitable to decimal GA is researched, on the basis of using for reference of the concept of gradient, an evolution direction based directed variant GA is proposed; 3.The mapping method is applied in the calculation of fitness function. Comparing with other methods the mapping method is simple and easy to use, no negative can appear in the calculation and a constant elective pressure can be applied to the group; 4.Independent mutation is applied to the optimal individual in each generation to improve the search efficiency. This unproved GA is used for optimizing the reactive power in Jinchang regional power network in Gansu province, and good results are obtained.
出处
《电网技术》
EI
CSCD
北大核心
2004年第10期42-44,54,共4页
Power System Technology
关键词
电力系统
地区电网
无功功率
优化
定向变异
遗传算法
Binary codes
Calculations
Electric network analysis
Electric power systems
Genetic algorithms
Integer programming
Mathematical techniques
Optimization
