A parallel multi-period optimal scheduling algorithm in microgrids with energy storage systems using decomposed inter-temporal constraints. (1st July 2020)

• Record Type: Journal Article
• Authors: Kim, Tae Hyun; Shin, Hansol; Kwag, Kyuhyeong; Kim, Wook
• Is Part Of: Energy. Volume 202(2020)
• Publication Date: 2020-07-01
• Subjects: Optimal generation scheduling -- Parallel computation -- Unit commitment -- Generalized Bender's decomposition -- Optimality condition decomposition; Power resources -- Periodicals; Power (Mechanics) -- Periodicals; Energy consumption -- Periodicals; 333.7905
• DOI: 10.1016/j.energy.2020.117669
• Languages: English
• ISSNs: 0360-5442

Abstract

Abstract: Because microgrids have relatively high share of renewable energy sources and energy storage systems (ESSs) compared with existing large-scale power systems, the inter-temporal constraints such as the generators' ramp-rates and the state-of-charge of the ESSs have a much greater impact on system operation. Therefore, in this paper, the optimization of the microgrid operation, the commitment of generators and the charging/discharging of ESSs, is formulated as a mixed-integer nonlinear programming (MINLP) problem with inter-temporal constraints. In order to find the optimal solution to the problem effectively, we propose a parallel computation method based on the generalized Bender's decomposition and the optimality condition decomposition. The method has the structure which is suitable for parallel computation and the convergence to the optimal solution is greatly improved compared with conventional sequential optimization methods. The proposed method is applied to the CIGRE medium-voltage microgrid benchmark system and the simulation results show that the proposed method has a potential for facilitating full-scale parallel computation ability and the application to the real-time operation of microgrid system. Highlights: Microgrid scheduling problem is solved via generalized Benders decomposition. Subproblem is further decomposed into the sub-subproblems. Each sub-subproblem is assigned to each core of the multi-core processors and solved simultaneously. The proposed parallel framework has been analyzed concerning the computation time and total cost.