Adaptive Recurrent Neural Network-Based ADRC Supplementary Damping Controller for SSO Mitigation in Type-3 Wind Power Systems

The integration of renewable energy sources impacts power system dynamics and increases the risk of sub-synchronous oscillations (SSO) in series-compensated doubly fed induction generators (DFIGs). The conventional subsynchronous damping controller (SSDC) is commonly used because of its ease of implementation. However, its effectiveness diminishes in wind- integrated systems characterized by nonlinear dynamics and significant operational uncertainties. To address this challenge, a supplementary damping controller utilizing a model-free control technique known as active disturbance rejection control (ADRC), based on a recurrent radial basis function neural network (RRBFNN) is proposed in this paper. The RRBFNN is employed to approximate the total disturbances resulting from parametric uncertainties, non-linear dynamics, and periodically varying disturbances. The utilization of RRBFNN aims to alleviate the constraints of conventional ADRC, specifically, its limited adaptability to manage the extended state observers (ESO). The stability of the closed-loop system with the proposed RRBFNN-ADRC is analyzed utilizing the Lyapunov method. The effectiveness of the proposed controller is evaluated using the 2009 Electric Reliability Council of Texas (ERCOT) event and an interconnected wind farm model. The controller hardware in loop (CHIL) experiments are conducted to evaluate the performance of the proposed controller in comparison to recently existing damping controllers under diverse operating conditions.