Adaptive Neural Network-Based Higher-Order Sliding Mode Control for Floating Offshore Wind Turbines

This paper introduces a novel adaptive feedback control approach for disturbed chains of integrators with smooth disturbances with unknown upper bound. The proposed approach combines adaptive neural network with higher-order sliding mode control to achieve the convergence of system states towards a vicinity of the origin. Notably, this approach does not rely on any prior information about the disturbance. The adaptive neural network term compensates the disturbance with an error, while the higher-order sliding mode control term effectively addresses this error and ensures the stabilization of the system state. Compared with existing neural network-based sliding mode control approaches, our proposed method does not require reducing the system order and utilizes only two terms for control. These characteristics contribute to its simplicity and lead to improved closed-loop performance. The effectiveness of the adaptive feedback control is specifically assessed for semi-submersible floating offshore wind turbines operating above rated speed. Simulation results demonstrate superior performance in rotor speed regulation and platform pitch reduction compared to the baseline gain-scheduling proportional integral controller.