Optimization of unit commitment and economic dispatch in microgrids based on genetic algorithm and mixed integer linear programming

• Published in: Applied energy Vol. 210; pp. 944 - 963
• Main Authors: Nemati, Mohsen, Braun, Martin, Tenbohlen, Stefan
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 15.01.2018
• Subjects: algorithms; Economic dispatch; electric power; energy; Genetic algorithm; issues and policy; linear programming; lithium batteries; Microgrids; Mixed integer linear programming; Unit commitment; Economic dispatch; Mixed integer linear programming; Genetic algorithm; Microgrids; Unit commitment
• ISSN: 0306-2619; 1872-9118
• DOI: 10.1016/j.apenergy.2017.07.007

•Day-ahead dispatching of the renewable energy resources inside a microgrid.•Genetic algorithm based optimizer for solving unit commitment and economic dispatch.•Aging model of the Li-Ion battery based on an event-driven method.•Mixed integer linear programming for optimal power flow of microgrids. Energy Management System (EMS) applications of modern power networks like microgrids have to respond to a number of stringent challenges due to current energy revolution. Optimal resource dispatch tasks must be handled with specific regard to the addition of new resource types and the adoption of novel modeling considerations. In addition, due to the comprehensive changes concerning the multi cell grid structure, new policies should be fulfilled via microgrids’ EMS. At the same time achieving a variety of (conflicting) goals in different microgrids requires a universal and a multi criteria optimization tool. Few of recent works in this area have considered the different perspectives of network operation with high amount of constraints and decision criteria. In this paper two dispatch-optimizers for a centralized EMS (CEMS) as a universal tool are introduced. An improved real-coded genetic algorithm and an enhanced mixed integer linear programming (MILP) based method have been developed to schedule the unit commitment and economic dispatch of microgrid units. In the proposed methods, network restrictions like voltages and equipment loadings and unit constraints have been considered. The adopted genetic algorithm features a highly flexible set of sub-functions, intelligent convergence behavior, as well as diversified searching approaches and penalty methods for constraint violations. Moreover, a novel method has been introduced to deal with the limitations of the MILP algorithm for handling the non-linear network topology constraints. A new aging model of a Lithium-Ion battery based on an event-driven aging behavior has been introduced. Ultimately, the developed GA-based and MILP-based optimizers have been applied to a test microgrid model under different operation policies, and the functionality of each method has been evaluated and compared together.