中文摘要：

根据梯级水电站优化调度特点,建立粒子群算法求解多阶段最优化问题数学模型。针对基本粒子群算法（PSO）在早期存在精度较低、易发散等缺点,后期出现＂趋同性＂和＂早熟＂等现象,提出了自适应多策略粒子群算法。并将该算法与基本PSO、改进型PSO、杂交PSO和收敛因子PSO分别在雅砻江梯级水库群优化调度中应用,通过对其优化结果的比较,验证了改进算法在加快收敛速度和提高算法精度方面的有效性。

英文摘要：

In this work a mathematical model is established to solve multi-stage optimization problem for the operation of cascade hydropower stations. The particle swarm optimization （ PSO） algorithm is of low accuracy,easy to diverge at the early stage,and tends to homoplasy and premature at the later stage. Thus an improved PSO or the adaptive multi-variant stratege PSO algorithm （AMPSO） is proposed. The optimization of cascade hydropowers on the Yalong river is solved separately by PSO,AMPSO,hybrid PSO and convergence- factor PSO. Results show an improvement of AMPSO in convergence and accuracy.