基于改进混合蛙跳算法的梯级水库优化调度

Optimizing Operation of Cascade Reservoirs Based on an Improved Shuffled Frog Leaping Algorithm

针对混合蛙跳算法在寻优过程中出现的早熟收敛问题,利用混沌技术的遍历性优势对子群最优个体进行变异操作,形成局部精细搜索策略;根据蛙群相对多样性参数来判断算法是否陷入局部最优,进而对蛙群最优个体进行扰动以提高全局寻优能力,形成全局激励调节策略。耦合2种策略,提出了一种改进混合蛙跳算法。将其应用于李仙江梯级水库优化调度中,结果表明所提算法具有寻优质量高、收敛速度快的特点,有效地克服了标准混合蛙跳算法的早熟缺陷,为水库调度模型的求解提供了一种新方法。

In view of the premature convergence in shuffled frog leaping algorithm（ SFLA）,an improved shuffled frog leaping algorithm（ AISFLA） is proposed by coupling the local refine search strategy（ LRSS） with the global incentive regulation strategy（ GIRS）. LRSS improves the local search ability by using chaos technology to conduct more refined search around the optimal individual of each group,while GIRS keeps an efficient global search performance by disturbing the optimum individual to improve the frogs＇ population diversity and further motivate frogs jumping out of stable state. AISFLA is applied to the optimal operation of Lixianjiang cascade reservoirs as a demonstration. The modeling result proves that AISFLA is of high optimization quality and fast convergence by effectively handling the premature convergence of SFLA,hence can be a new approach to the solution of optimal operation of cascade reservoirs.