Research on Adaptive PI Controller of LCL-Filtered Voltage Source Rectifier with Neural-Network

In this paper, a neural-network-based adaptive PI current controller of voltage source rectifier (VSR) is proposed. Complex vector PI decoupling scheme can eliminate the voltage coupling completely, and the neural network algorithm can tune the parameters of current regulator online. The novel controller has a fast dynamic response, strong adaptability to those occasions such as the nonlinear model and variable load. A double loop control strategy with an outer dc link voltage control based on traditional PI controller and inner current control based on the novel controller is discussed in this paper. Simulation results verify the correctness and feasibility of the neural-network-based PI controller.