基于时空组合动态规划的分解协调模型在并联站群优化中的运用  

Application of decomposition-coordination model on dynamically programmed space-time combination to operation optimization of parallel pumping stations

作　　者：仇锦先[1] 程吉林[1] 张仁田[2] 张礼华[1] 

机构地区：[1]扬州大学水利科学与工程学院,扬州225009 [2]江苏省水利勘测设计研究院有限公司,扬州225009

出　　处：《水力发电学报》2011年第6期183-188,共6页Journal of Hydroelectric Engineering

基　　金："十一五"国家科技支撑计划重大项目(2006BAB04A03);国家自然科学基金项目(60974099);江苏省属高校自然科学面上项目(11KJB57002)

摘　　要：并联站群叶片全调节变频变速组合优化属于多单元、多机组、多时段、多决策变量复杂非线性系统优化问题,本文引入基于时空组合动态规划的分解协调模型,即协调层以空间各子系统为阶段,提水总量为状态变量,各阶段水量分配系数为决策变量,采用动态规划法进行各子系统间的总体水量协调;分解层各子系统以时间为阶段,各阶段开机台数、叶片角度及转速为决策变量,分配水量为状态变量,采用动态规划逐次渐近法确定最优开机台数及运行方式。通过在江都站并联站群组合优化中的应用,确定了各站各时段开机台数、叶片安放角度及机组转速的最优组合方案,实现了系统运行机组耗电费用最小时的优化调度过程,表明了方法采用的可行性。Operation optimization of parallel pumping stations featured by blade-adjusting and speed-varying pumps is a nonlinear problem of complex multiple decision variables.This paper presents a decomposition-coordination method of dynamically programmed space-time combination.In this method,spatial subsystems are used as coordination layer＇s stages and water distribution coefficients at each stage as decision variables,and dynamic programming is adopted for overall coordination of the subsystems＇ water quantities.Further,on each time-staged decomposition layer the decision variables of operating-pumps number,blade angles and rotational speeds are formulated,and successive approximation of dynamic programming is used to determine the optimal operation mode of every subsystem.Application of the method to the blade-adjusting and speed-varying Jiangdu pumping station shows feasible results in optimizing these decision variables.

关 键 词：水力机械 变角变速组合优化 分解协调 时空组合动态规划 并联站群 

分 类 号：TV737[水利工程—水利水电工程]