论文部分内容阅读
本文介绍了应用多维随机动态规划解决贵州三库七水电站梯级发电系统的联合优化调度的实例。根据实例的具体条件,对其中两个梯级水库制定了“同步蓄放”的拟优原则,使三库问题概化为两库的数学模型。文中对库群调度中某个水厍受到极限水位的制约不能按指示决策工作的情况,首次提出了“补偿调度”的原则。与此同时,对相应的寻优方法也作了改进。这些都在数学模型中得到了应用和检验。与时历操作结果相比,优化效益十分明显,说明模型的概化与应用是成功的。
This paper presents an example of applying multi-dimensional stochastic dynamic programming to solve the joint optimal operation of cascade power generation system in San Kuqi hydropower station. According to the specific conditions of the examples, the principle of “simultaneous storage” was formulated for the two cascaded reservoirs, and the three-storey problem was generalized into the mathematical model of the two reservoirs. In this paper, the principle of “compensation schedule” is first put forward for the first time when the restriction of a water level in a reservoir group dispatched by the limit water level can not be decided by instruction. At the same time, the corresponding optimization method has also been improved. These are applied and tested in mathematical models. Compared with the results of the calendar operation, the optimization benefit is very obvious, which shows that the generalization and application of the model are successful.