自优化模拟及其在南水北调东线工程中的应用

Self-optimization simulation and its application in the Project of South-to-North Water Transfer

本文在常规的模拟技术基础上,增加在线辨识、控制等环节、对模拟过程施行自适应目标控制,提出了一种多水库联合最优调度的自优化模拟模型。并针对南水北调东线一期工程五湖泊大型复杂系统,研究了时、空顺序不同的模拟过程的功能特点,采用顺、逆时序和顺、逆流向结合的模拟控制技术以及交替放松约束并随后收紧的措施,处理决策控制的后效性,提出了“逐次收紧约束自适应目标控制模拟技术”,在实践中成功地实现了模拟模型自行优化的目的。在实例计算中,解决了五个湖泊联合运行的优化调度问题,证明了自优化模拟模型具有仿真性好、收敛速度快、计算时间仅随水库数目线性增加的优点,开辟了一条多水库系统最优调度的新途径。

A new self-optimization simulation model is developed for optimum operation of multireservoir systems by adding the feed-back elements of on-line identification and control into a common simulation model for self-adaptive control. Based on studying different sequences of simulating derection in time and space, considering time-forwards and time-backwards as well as the main current and the adverse current sequances with measures of losing and lightening up constraints alternatirely, a self-adaptive control simulation technique is presented. It is satisfactorily used in the optimum operation study for the Project of South-to-North Water Transfer which is a large, complex water resources system with five lakes. It has been indicated that the self-optimization simulation model behaves better in imitation and convergence as compared with other mathematical models. The CPU time cost increases linearly with the number of reservoirs.