遗传算法在电力系统经济负荷分配中的应用  

Application of Genetic Algorithm on Economic Dispatch of Power Systems

作　　者：何大阔[1] 王福利[1] 毛志忠[1] 李鸿儒[1] 

机构地区：[1]东北大学教育部暨辽宁省流程工业综合自动化重点实验室,辽宁沈阳110004

出　　处：《系统仿真学报》2007年第4期890-892,900,共4页Journal of System Simulation

基　　金：国家自然科学基金资助项目(60674063;60374003);973子课题资助项目(2002CB312200)

摘　　要：针对考虑阀点效应的电力系统经济负荷分配问题的特点,将遗传算法与传统基于梯度的数学优化方法相结合,提出一种求解电力系统经济负荷分配问题的混合遗传算法。通过极大熵理论将经济负荷分配问题近似为可导问题,并将SQP法引入遗传算法提出SQP算子,以提高遗传算法的寻优速度与局部搜索能力。基于优化设计理论应用均匀设计确定初始种群。同时,利用单纯形交叉算子的多点交叉特性使种群逐步趋近全局最优点。实例研究结果验证了方法的有效性。According to the characteristics of economic dispatch （ED） problem with value-point effects of power systems, genetic algorithm and conventional optimization method based on gradient were associated, A hybrid genetic algorithm was proposed to solve economic dispatch （ED） problem of power systems. Economic dispatch （ED） problem was approximated by a differentiable one using a maximum entropy method. SQP operator was proposed to improve the searching speed and the local searching capability of genetic algorithm. Based on the optimization design theory, the initial population was established by using uniform design. At the same time, simplex crossover operator was applied to lead the population to the global optimum. The study on case proves the validity of this algorithm.

关 键 词：经济负荷分配 混合遗传算法 极大熵 SQP算子 单纯形交叉算子 

分 类 号：TP13[自动化与计算机技术—控制理论与控制工程]