混沌变异粒子群算法在水库供水调度规则制定中的应用

本文针对基于调度图规则的水库供水调度问题,建立了以水库供水保证率高且缺水量少为目标的优化调度模型。同时应用混沌变异减缓粒子群算法收敛速度,当算法进化停滞步数大于停滞步数阀值时,随机选取其中20%的粒子进行混沌变异操作,将原本聚集的粒子群"驱散开来",达到增加种群多样性、避免算法早熟收敛的目的,并将该算法引入到调度图的获取中。并以白石水库为例,得到了满足各项用水保证率的水库调度图,验证了该方法的可行性。

Application of chaotic variation particle swarm optimization in reservoir water supply dispatching rules making

In the paper,optimization dispatching model with high reservoir water supply guarantee rate and less water deficiency variation as targets is established aiming at reservoir water supply dispatching problems based on dispatching chart rules.Meanwhile,the convergence rate of particle swarm optimization algorithm is slowed down by applying chaotic variation.When algorithm evolutionary stagnation steps are greater than stagnation step threshold,20% particles thereof are randomly selected for chaotic variation operation.The originally gathered particle swarms are ‘dispersed’to reach the purposes of increasing population diversity and avoiding premature convergence.The algorithm is introduced into the dispatching chart acquisition.Baishi Reservoir is adopted as an example for obtaining the reservoir dispatching chart meeting all water consumption guarantee rates.The feasibility of the method is verified.