基于分段粒子群算法的梯级水库多目标优化调度模型研究

Multi-objective optimal dispatch model of cascade reservoirs based on piecewise particle swarm optimization algorithm

提出了约束破坏向量、分段粒子群算法以及多目标分段粒子群算法,有效解决了在时间步长较小、计算时段数目较多时,传统智能优化算法解水库优化调度问题的寻优效率低下甚至无可行解的问题。该方法基于粒子群算法框架,引入约束破坏向量、分段操作和特殊变异操作来增强进化过程中的种群质量,从而提高算法的计算效率。闽江流域金溪梯级水库多目标优化调度的实例分析表明,在解时间步长较小、计算时段数目较多的水库优化调度问题时,分段粒子群算法、多目标分段粒子群算法相对其他算法具有明显优势。

Conventional intelligent optimization algorithms are inefficient or even impossible in finding feasible solutions to the reservoir optimal dispatch calculations with small time step and large number of calculation periods.On this basis,the damaged vector constraint,piecewise particle swarm optimization(PPSO)algorithm and multi-objective piecewise particle swarm optimization(MOPPSO)algorithm are proposed to successfully solve the problem.This method introduces the damaged vector constraints,piecewise operation and special mutation operation to enhance the quality of populations in the evolution process,thus improving the computational efficiency.Then the algorithms were applied to the multi-objective optimal dispatch of Jinxi cascade reservoirs in Minjiang River Basin.The results show that PPSO and MOPPSO have obvious advantages over other algorithms in the optimal dispatch of reservoirs with small time step and large number of calculation periods.