基于协同进化遗传算法的水库群供水优化调度研究

Research on Co-Evolutionary Genetic Algorithm for Reservoir Optimization Water Supply Dispatching

针对水库群供水优化调度问题,介绍了一种改进的协同进化遗传算法。该算法针对求解高维、复杂的水库群优化调度时多约束条件难以处理、计算机时长、易陷入局部最优解等缺陷,建立了相应的罚因子的评价机制,生成了两类进化子种群,运用改进遗传算法同时对不同种群进行操作,并将其应用在滦河下游六水库联合供水优化调度中。实例计算结果表明,用该算法求解水库群供水优化调度问题,结果可靠、合理,计算效率高。

Aiming at the problems of reservoir optimization water supply dispatching,an improved co-evolutionary genetic algorithm is described.For high-dimensional and complex cascade reservoirs,the traditional optimization algorithms are difficult to deal with multi-constraint condition、a long computing time、falling into local optimal solution easily and other defects,a corresponding penalty factor of the evaluation mechanism is established.Two evolution subpopulations are formed,and at the meantime,the improved genetic algorithm is applied to operate the various subpopulations.It is applied to six reservoirs to optimize water supply dispatching in the lower reaches of the Luan-he river.The results from the real example calculation indicate that when this algorithm is used to solve the problem of the optimization dispatching of reservoir water supply,the results are reliable and rational with high calculation efficiency.