A Sliding Mode Control Strategy Based on RBF Neural Network for Grid-connected Inverter

To address the operational challenges of the grid-connected inverter under parameter deviations, grid impedance fluctuations, and unmodeled disturbances, this paper proposes a sliding mode control method based on the Radial Basis Function Neural Network (RBFNN). Initially, a sliding mode surface is designed to obtain the equivalent control law corresponding to the perturbed system; Subsequently, a switching control term is given to accelerate the dynamic response capability of the system. The RBFNN is then employed to estimate the aggregated uncertainties through online approximation, enabling the synthesis of a composite control law. Lyapunov stability analysis rigorously demonstrates the system's asymptotic convergence. Matlab/Simulink simulation results demonstrate that under parameter perturbations, time-varying grid impedance, and other unmodeled disturbances, the proposed method exhibits significantly reduced approximation errors and faster response speed, effectively enhancing the robustness of the system and ensuring stable operation of the grid-connected inverter system.