论文部分内容阅读
由贝尔曼(Bellman,1957)提出的动态规划是一种多阶段决策过程的优化方法。这种方法能广泛应用于水资源系统规划和管理中的各种决策问题,这主要是由于水资源系统中的大量问题具有非线性和随机性的特征。用动态规划求解水库调度问题,一般将蓄水量作为状态变量,放水量作为决策变量,用水量平衡方程作为状态转移方程,并按时间顺序划分为阶段,根据所求目标及贝尔曼最优原理写出递推方程。这样将具有最大量变量的复杂问题分阶段划为一系列简单的小问题用递推方法求解。尤其当目标函数带有非线性问题(例如:发电),并考虑水库入流的随机性,利用动态规划更显示出它的优越性。
The dynamic programming proposed by Bellman (1957) is an optimization method of multi-stage decision process. This approach can be widely applied to various decision-making problems in the planning and management of water resources systems, mainly due to the non-linear and random nature of a large number of problems in water resources systems. The dynamic programming is used to solve the reservoir scheduling problem. Generally, the water storage is taken as the state variable and the water discharge is taken as the decision variable. The water balance equation is used as the state transfer equation, which is divided into phases according to the time sequence. According to the target and Berman’s optimal principle Write a recurrence equation. In this way, the complex problem with the largest amount of variables is divided into a series of simple and small problems by recursion method. Especially when the objective function has non-linear problems (for example, power generation) and considering the randomness of reservoir inflow, the dynamic programming also shows its superiority.