A MPC-based load frequency control considering wind power intelligent forecasting

Abstract

Currently, the significant randomness of wind power hampers the stability of the grid system. Furthermore, existing control strategies, which solely rely on current measured wind power output, are inadequate in addressing the rapid and continuous fluctuations of wind power. Based on this, a novel load frequency control (LFC) that combines wind power prediction and model predictive control (MPC) is proposed in this paper. The high-precision wind power forecasts are embedded into the MPC, enabling MPC to develop robust control strategies that flexibly respond to the random variability of wind power. For wind power prediction, an improved Reformer model with inversion and gated linear unit (GiReformer) is constructed, which achieves multi-step predictions of wind power at the microscale. In addition, Laguerre function is introduced in MPC to reduce the computational load, and the settings for frequency constraints, generate rate constraints (GRC), control input constraints and terminal constraints ensures the safe and stable operation of the power grid. According to simulations in a high-proportion hydropower system and a multi-energy and multi-regional interconnected power system, the proposed method alleviates system frequency fluctuations up to 71.88 % and 51.78 %, respectively, compared to the comparative methods. In addition, constraints are well handled by the proposed method.