Optimal scheduling of a renewable micro-grid in an isolated load area using mixed-integer linear programming

• Published in: Renewable energy Vol. 35; no. 1; pp. 151 - 156
• Main Authors: Morais, Hugo, Kádár, Péter, Faria, Pedro, Vale, Zita A., Khodr, H.M.
• Format: Journal Article
• Language: English
• Published: Oxford Elsevier Ltd 2010; Elsevier
• Subjects: Applied sciences; Economic data; Electric energy; Energy; Energy economics; Exact sciences and technology; General, economic and professional studies; Methodology. Modelling; Mixed-integer linear programming; Natural energy; Renewable energy park; Scheduling; Virtual power producer; Hungary; Europe; Hungary, Budapest; Virtual power producer; Renewable energy park; Scheduling; Mixed-integer linear programming; Isolated power system; Linear programming; Mixed integer programming; Wind generator; Constrained optimization; Case study; Optimal strategy; Renewable energy; Decentralized system; Battery; Photovoltaic array; Electric power production; Photovoltaic generator; Energy management; Fuel cell; Energy storage
• ISSN: 0960-1481; 1879-0682
• DOI: 10.1016/j.renene.2009.02.031

In the energy management of the isolated operation of small power system, the economic scheduling of the generation units is a crucial problem. Applying right timing can maximize the performance of the supply. The optimal operation of a wind turbine, a solar unit, a fuel cell and a storage battery is searched by a mixed-integer linear programming implemented in General Algebraic Modeling Systems (GAMS). A Virtual Power Producer (VPP) can optimal operate the generation units, assured the good functioning of equipment, including the maintenance, operation cost and the generation measurement and control. A central control at system allows a VPP to manage the optimal generation and their load control. The application of methodology to a real case study in Budapest Tech, demonstrates the effectiveness of this method to solve the optimal isolated dispatch of the DC micro-grid renewable energy park. The problem has been converged in 0.09 s and 30 iterations.