A Layered Composite Control Strategy for Voltage Source PWM Rectifiers Based on Improved Sliding Mode and Back‐Propagation Neural Network PID

To address the poor dynamic response and limited disturbance rejection of conventional dual‐loop PI control in three‐phase voltage source PWM rectifiers (VSRs), this paper proposes a composite control strategy featuring an improved sliding mode control (SMC) for the voltage outer loop and a back‐propagation neural network (BP‐PID) controller for the current inner loop. First, a mathematical model with feedforward decoupling is established in the synchronous rotating (dq) frame. For the DC‐side voltage outer loop, an improved SMC is designed by integrating an integral term into the sliding surface. This enhancement mitigates the slow response and steady‐state error of conventional SMC under high‐frequency disturbances and parameter variations. For the current inner loop, a BP‐PID adaptive controller is introduced to compensate for nonlinearities and improve current tracking precision, thereby reducing the impact of current fluctuations on the DC‐link voltage. Finally, simulations are conducted in MATLAB/SIMULINK under step changes in voltage reference and load variations. The results verify that the proposed control strategy reduces the recovery time and overshoot of the DC voltage, significantly improving both the dynamic response and the robustness against disturbances.