【摘要】 针对传统遗传算法存在早熟及收敛速度慢等问题,采用了改进的遗传算法和局部搜索能力很强的单纯形法相结合的方法进行负荷模型的参数辨识。具体研究了算法的软件实现,包括编码、初始种群形成、适应度函数确定、自适应交叉变异、移民策略、排序选择、单纯形优化及终止条件的实现方法。算例表明了经改进后的混合遗传算法具有辨识精度高、速度快和参数分散性小等优点。更多还原

【Abstract】 The traditional genetic algorithm has defects,including early mutating and slow convergence speed.The improved genetic algorithm and simplex algorithm are adopted to identify parameters of the load model.The software is developed,including coding,format of initial population,fitness fuction,adaptive crossover and mutation,immigration trategy,ranking selection,simplex algorithm and termination condition.Case study shows that the improved hybrid genetic algorithm is better than the traditional one,which is embodied in fast convergence speed,high precision and low dispersion.更多还原