Solution for short-term hydrothermal scheduling with a logarithmic size mixed-integer linear programming formulation. (15th March 2019)

• Record Type: Journal Article
• Authors: Jian, Jinbao; Pan, Shanshan; Yang, Linfeng
• Is Part Of: Energy. Volume 171(2019)
• Page Start: 770
• Page End: 784
• Publication Date: 2019-03-15
• Subjects: Short-term hydrothermal scheduling -- Piecewise linearize -- Logarithmic size -- Mixed-integer linear programming -- Non-linear programming; Power resources -- Periodicals; Power (Mechanics) -- Periodicals; Energy consumption -- Periodicals; 333.7905
• DOI: 10.1016/j.energy.2019.01.038
• Languages: English
• ISSNs: 0360-5442

Abstract

Abstract: Short-term hydrothermal scheduling issue is usually hard to tackle on account of its highly non-convex and non-differentiable characteristics. A popular strategy for handling these difficulties is to reformulate the issue by various linearization techniques. However, in this process, a fairly large number of continuous/binary variables and constraints will be introduced, which may result in a heavy computational burden. In this study, a logarithmic size mixed-integer linear programming formulation is presented for this issue, that is, only a logarithmic size of binary variables and constraints will be required to piecewise linearize the nonlinear functions. Based on such a formulation, a global optimum is therefore can be solved efficiently. To remove the linearization errors and cope with the network loss, a derivable non-linear programming formulation is derived. By optimizing this formulation via the powerful interior point method, where the previous global solution of mixed-integer linear programming formulation is used as the starting point, a promising feasible solution is consequently attained. Numerical results show that the presented logarithmic size mixed-integer linear programming formulation is more efficient than the generalized one and when it is incorporated into the solution procedure, the proposed methodology is competitive with currently state-of-the-art approaches. Highlights: A logarithmic size mixed-integer linear programming formulation is presented. The introduced binary variables and constraints are dramatically reduced. A solution methodology is developed for short-term hydrothermal scheduling. Numerical results show the effectiveness of the proposed model and methodology.