以发电量和保证出力为目标建立梯级水电站的多目标发电优化调度模型,对三峡梯级中长期发电优化调度进行研究。针对传统方法求解多目标优化问题的局限,提出一种强度Pareto差分进化算法(Strength Pareto Differential Evolution,SPDE)用于求解梯级水电站的多目标发电优化调度问题。SPDE以差分进化算法(Differential Evolution,DE)为基础,采用SPEA2的适应度评价方法,并根据多目标优化的特点对DE的进化算子进行修正。同时,提出一种自适应柯西变异策略(Adaptive Cauchy Mutation,ACM)用于克服算法的早熟收敛问题。三峡梯级水电站实例研究结果表明,SPDE可同时考虑两个目标并有效处理复杂约束条件,一次运行即可得到一组在各目标分布均匀、分布范围广的非劣调度方案供决策者评价优选。
A multi-objective optimization model for operating cascade hydropower stations is established.The model can maximize both firm power and annual energy production simultaneously.The model is solved using the strength Pareto differential evolution(SPDE) approach.The latter is a novel extension of the Differential Evolution(DE) algorithm to multi-objective optimization problems.SPDE adopts the fitness assign method used in the strength Pareto evolution approach 2(SPEA2),and modifies the DE operators according to the characteristic of multi-objective optimization.An adaptive Cauchy mutation(ACM) is also used to prevent premature convergence.The model is tested using a case of Three Gorges cascade hydropower stations.The result shows that SPDE is able to simultaneously consider the two optimization objectives and effectively deal with complex constraints.A set of alternative non-dominated schemes with uniform coverage can be provided for decision making,which presents a viable alternative for multi-objective optimal dispatch of cascade hydropower stations.Our result could be used as a reference for making the mid-long term generation scheduling schemes for Three Gorges cascade hydropower stations.