梯级水库联合优化调度的差分演化算法研究

Discussion on the application of differential evolution algorithm for optimal operation of cascade reservoirs

差分演化算法是一种收敛快、控制参数少并且易于实现的演化算法。本文在探讨梯级水库联合优化调度的差分演化算法中,针对传统优化算法在高维梯级水库优化调度时多约束条件难以处理、计算机时长等缺陷,提出了非参数惩罚的约束条件处理机制,并对差分演化算法进行了改进。通过实际算例验证了该方法的可靠性,从而为高维梯级水库的优化调度提供了一种新的求解途径。

A differential evolution algorithm implemented. It is introduced and applied to penalty is fastly convergent only with few control the operation of cascade reservoirs in this parameters and simply paper. A non-parameter constraint handling methods is presented to improve the problem of handling more restricting conditions and long computing time compared with tradition intelligence optimum algorithm applied to optimal operation of cascade reservoirs, and the differential evolution algorithm is improved. The credibility of the algorithm is verified by the practical example, and it provides a new solving way for the cascade reservoirs optimal operation.