A model-free adaptive controller with tracking error differential for collective pitching of wind turbines

Abstract

Abstract The aerodynamic power captured by wind turbines is generally limited by pitch control when it operates above rated wind speed. However, problems such as actuator constraints, unmeasured system states, and stochastic wind speed fluctuations, make it difficult to establish an accurate model of the controlled collective pitch system. This paper proposes an improved model-free adaptive controller (MFAC) by considering the differential of tracking error in cost function (MFAC with the differential of tracking error, DE-MFAC), to compensate for potential effect of the system dynamic characteristics under random disturbance. Meanwhile, a new parameter β is introduced to adjust the impact of the compensation module to alleviate the contradiction between rapidity and hypertonicity. Then, the bounded-input bounded-output (BIBO) stability and monotonic convergence of the tracking error are derived by the contraction mapping. Furthermore, a Simulink/FAST platform is established and simulated under various wind conditions. The results show that, compared with the baseline PI and the MFAC based on compact form dynamic linearization (CFDL-MFAC), the proposed DE-MFAC is capable of handling the time-delay and large inertia of collective pitch systems, which not only significantly makes improvements in robustness and dynamic response, but also limits the average fluctuating amplitude of generator power within 1%.