摘要
遗传算法的主要的特点是:简单、通用、鲁棒性强,能依概率收敛到问题的全局最优解。用遗传算法来求解多目标的电力网络的优化规划,可以避免局部最优的问题。针对现有的遗传算法的不足和对初始种群的依赖,提出了改进的遗传算法,采用贪心法进行种群初始化,从而避免了遗传算法初始种群过于随机,提高了算法的收敛速度和搜索性能。
Genetic algorithm （GA） is not only simple, universal and robust, but also can obtain the global optimal solution based on probability. GA can be used to solve the multi-objective power grid planning and avoid the local optimization problem. Given the disadvantages of existing genetic algorithm and the dependence on the original population, an improved Genetic Algorithm is proposed, in which greedy method is adopted for producing original population in order to avoid randomness of the initial population of GA and increase the convergence speed and searching capability of GA.
出处
《华东电力》
北大核心
2013年第10期2095-2098,共4页
East China Electric Power
关键词
遗传算法
电网规划
贪心法
genetic algorithm
grid planning
greedy method
