A Stochastic Model Predictive Control-Based Energy Management Approach for Microgrids With Electric Vehicles
Weilin Yang, Haojie Fang, Dezhi Xu, Bin Jiang, Peng Shi
Abstract
A two-layer stochastic model predictive control (MPC) framework is proposed in this article to address uncertainties associated with electric vehicles (EVs) in microgrid energy management. At the upper layer, a multiscenario MPC (MS-MPC) approach is considered, which utilizes multiscenario sampling to address the inherent uncertainties with respect to the number of available EVs and their states of charge. The proposed approach effectively minimizes prediction errors while ensuring the cost-efficient operation of the energy management system (MES). To mitigate concerns related to increased convergence time due to scenario combinations, a two-stage scenario reduction technique is introduced to improve the computational efficiency. At the lower layer, an EV aggregator allocation model is developed that considers the power requirements of each individual EV for both charging and discharging operations. Simulation results demonstrate the effectiveness of the proposed MS-MPC approach, revealing its high cost-effectiveness performance and affordable computational burden.