New Planning Framework of Hydro-thermal System with Annual Water Consumption Constraints

This paper studies the planning problem of hydro-thermal system with annual water consumption constraints, which are motivated by the practical system in the Yellow River Basin of China. This problem is formulated as one multistage stochastic optimization problem. The high-dimensional, dynamic, nonlinear and stochastic characteristics of hydro-thermal systems are considered. Since we can hardly get the accurate estimation of natural inflows in a further future, a new framework, the-closer-the-more framework, is proposed for scenario tree construction. This framework can help to reduce the size of the problem. The solution approach is nonlinear stochastic dual dynamic programming, which is based on the approximation of the expected-cost-to-go function of stochastic dynamic programming by a piecewise linear function. Moving window method is used for planning. At each stage, annual water consumption constraints are divided two parts in optimization problem, one is for current year and the other is for next year. Numerical results are also presented for illustration. Efficiency and practicality can be seen in our simulation on several situations. 