Tube-Based Stochastic Model Predictive Control With Application to Wind Energy Conversion System
Xiangjie Liu, Le Feng, Xiaobing Kong, Shifan Guo, Kwang Y. Lee
Abstract
Model predictive control (MPC) is a powerful tool for modern wind energy conversion system (WECS). However, the stochastic uncertainty existed in wind speed presents a great challenge for the MPC to realize the optimal control task and guarantee the system stability. In this aspect, a stochastic MPC strategy with probabilistic tube and robust tube is proposed to realize the high-precision output power tracking with a scarce probability of exceeding the rated value. The actual state is restrained within the robust tube centered on the probabilistic tube with recursive feasibility by minimizing state error, while the nominal state is steered toward the rated point by designing a probabilistic tube based on the nominal stochastic model. Therefore, the stochastic wind speed disturbance is effectively rejected by the proposed tube-based stochastic MPC strategy so that the security of WECS operating in high wind speed regions can be guaranteed. A series of virtual plant simulations, including the experiments on National Renewable Energy Laboratory (NREL) fatigue, aerodynamics, structures, and turbulence (FAST), are carried out to validate the effectiveness of the proposed controller under different scenarios.