Model-Free Predictive Current Control With Robust Reference for Grid-Connected Voltage Source Inverter With LCL Filter

Model predictive current control often suffers from limited robustness due to its reliance on accurate model parameters of inductor-capacitor-inductor (LCL) filtered voltage source inverters (VSIs). To solve this problem, a model-free predictive current control with robust reference is proposed for LCL filtered VSIs. First, a triple-vector model-free prediction is proposed based on system variable gradients for reducing current ripple without requiring model parameters. Next, a two-step updating mechanism is established to refresh the system variable gradients, so as to enhance the accuracy of the system variable gradients and triple-vector model-free prediction. In addition, a robust reference is generated by combining sampled data and mathematical models, which mitigates the effects of parameter variations on reference accuracy. Finally, both simulation and experiment demonstrate the effectiveness of the proposed method.