基于遗传算法的梯级水库多目标联合调度仿真

Multi-Objective Joint Operation Simulation of Cascade Reservoirs Based on Genetic Algorithm

针对传统调度方法存在年调节保证出力不稳定、水位不平稳等问题,提出基于遗传算法的梯级水库多目标联合调度方法。以多目标发电联合调度和多目标防洪联合调度为主目标函数,以计算周期时间段内的发电总量最大、供水量最大、汛期阶段的弃水量最小等为子目标函数,以电站下的泄流量、电站出力、水库水量上下限、河道演进等为目标函数的约束条件,实现梯级水库多目标联合调度模型构建。引入遗传算法,设置种群数量等算法初始化参数,获取个体电站出力等一系列信息数据。通过种群之间适应度信息与个体的距离信息维持种群多样性,进行最优代数选取,将所得最优解作为满足梯级水库多目标联合调度条件的方案。对比传统方法实验结果表明,研究方法的年调度保证出力平稳性高,可有效实现发电与防洪等方面的调度,具为梯级水库多目标联合调度提供有利依据。

In traditional methods,firm power and water level are unstable.Therefore,a multi-objective joint scheduling method for cascade reservoir based on genetic algorithm was proposed.The multi-objective power joint scheduling and multi-objective flood control scheduling were taken as the objective functions.The maximum generating volume within each cycle,the maximum water supply,the minimum surplus water in flood season were taken as sub-objective functions.The discharging capacity under power station,the firm power,upper and lower limits of reservoir water amount and river channel evolution were taken as the constraint conditions.Thus,the construction joint scheduling model of cascade reservoirs was completed.The genetic algorithm was introduced,and the initial parameters of algorithm such as the population number were set.After that,a series of information data such as the firm power of individual power station were obtained.The fitness between populations and the distance between individuals were used to maintain the diversity of population.After the optimal algebraic was selected,the optimal solution was taken as a scheme to meet the condition of multi-objective joint scheduling of cascade reservoir.Simulation results show that the proposed method can ensure high stability of annual regulation firm power and effectively realize the joint scheduling of power generation and flood control.This method provides a favorable basis for multi-objective joint scheduling of cascade reservoir.