一种改进的加速遗传算法及其在机组动态负荷分配中的应用  

An improved Accelerating Genetic Algorithm and Its Application in Dynamic Load Distribution of Units

作　　者：方彦军[1] 崔春雷[1] 

机构地区：[1]武汉大学自动化系,武汉430072

出　　处：《华东电力》2010年第5期717-720,共4页East China Electric Power

基　　金：国家自然科学基金项目(60772107)

摘　　要：为解决遗传算法的收敛速度和全局收敛性之间的矛盾,基于对遗传算法收敛速度的分析,提出了一种改进的加速收敛的遗传算法,即基于双倍体初始种群、排序选择、自适应遗传参数和压缩空间法的实数编码加速遗传算法(AGA)。将其应用于实时动态负荷优化分配问题中,仿真结果表明对于复杂的多变量多约束问题,AGA具有较快的收敛速度,且能够收敛到全局最优解,能在一定程度上满足动态负荷优化分配对算法运算速度的要求。To solve the contradiction between the convergence speed and the global convergence of the genetic algorithm,an improved accelerating genetic algorithm（AGA）,i.e.real coding AGA with diploid initial population,ranking selection,adaptive genetic parameter and compressed space method,was proposed based on analyzing the convergence speed of the genetic algorithm.The improved algorithm was applied in the real-time dynamic load optimal distribution.The Simulation results showed that for the complex multi-variable and multi-constrained problems,AGA had a faster convergence speed and could converge to the global solution and meet the requirements of dynamic load optimal distribution for the computation speed to certain extent.

关 键 词：加速遗传算法 自适应参数法 压缩空间法 动态负荷优化分配 

分 类 号：TP29[自动化与计算机技术—检测技术与自动化装置]