GPU-Based Enumeration Model Predictive Control of Pumped Storage to Enhance Operational Flexibility

Abstract

With the integration of more renewable energy, operational flexibility becomes a bottleneck for power system operation. Pumped storage units can enhance their operational flexibility by switching between different operation modes. They have limited capacity for adjustment when operated as generators and are non-adjustable when operated in the pump mode. However, the power ramps generated during the switching of operation modes are valuable for regulating the frequency and increasing the operational flexibility. In order to achieve sufficient performance, it is essential to determine the optimal switching time for the pumped storage units. However, there is no standard method to obtain the optimal switching time due to the complexities and non-linear characteristics of such a problem. In this paper, an enumeration based model predictive control (MPC) strategy is proposed to determine the optimal switching time of a pumped storage unit to enhance its operational flexibility and facilitate frequency regulation. Furthermore, a graphics processing unit accelerated computing method is proposed to solve the problem effectively and to make the MPC controller suitable for practical applications.