基于差分进化‑自适应Metropolis算法的嘉陵江水库群联合供水调度  

Joint water supply dispatching of Jialing River reservoir group based on differential evolution‑adaptive Metropolis algorithm

作　　者：程毅 CHENG Yi(Evaluating and Examining Center of State‑Funded Construction Projects,Beijing 100037,China)

机构地区：[1]国家发展和改革委员会国家投资项目评审中心,北京100037

出　　处：《武汉大学学报（工学版）》2024年第9期1232-1240,共9页Engineering Journal of Wuhan University

摘　　要：差分进化-自适应Metropolis(differential evolution-adaptive Metropolis,DEAM)算法满足各态遍历性,且善于处理多变量非线性、高维、多峰值等情况,是一种新颖的全局优化求解方法。阐述了DEAM算法的求解机制,并将其应用于水库群优化调度。针对嘉陵江水库群的联合供水调度过程建立了多目标优化模型,基于水库群供水调度规则评价指标体系和多属性决策,最终得到最优的供水调度方案。结果表明:宝珠寺水库在枯水年能完全满足规划供水需求,而升钟、亭子口2座水库则需要对供水调度评价指标进行权衡。研究成果为DEAM算法应用于水库群优化调度研究奠定了一定的理论基础,可为制定嘉陵江水库群实际联合调度方案提供参考。The differential evolution-adaptive Metropolis(DEAM)algorithm,a novel global optimization method,satisfies the ergodicity of each state,and is good at dealing with multi-variable nonlinear,high-dimensional,multi-peak and so on.This paper introduces the solution mechanism of the DEAM algorithm,and subsequently applies this algorithm to the optimal operation of a reservoir group.A multi-objective optimization model is established for the joint water supply dispatching process of Jialing River reservoir group,and the optimal water supply dispatching scheme is finally obtained based on the evaluation index system and multi-attribute decision-making of water supply dispatching rules of the reservoir group.The results show that Baozhusi Reservoir can fully satisfy the demand of planned water supply in dry year,while Shengzhong and Tingzikou reservoirs need to weigh the evaluation index of water supply dispatching.The research results lay a theoretical foundation for the application of DEAM algorithm in the study of reservoir group optimal operation,and can provide a reference for the formulation of practical joint scheduling scheme of Jialing River reservoir group.

关 键 词：差分进化 自适应Metropolis 嘉陵江水库群 联合供水调度 

分 类 号：TV697.1[水利工程—水利水电工程]