Research on the implementation of a novel switching converter´s controller

The DC-DC buck converter is token as the controlled object, and a neural network controller is designed for this converter, hardware implementation of the controller is discussed based on FPGA. The reconfigurable feature of the the neural network controller is analyzed, several reconfigurable unit is given, the implementation model of this controller is set up based on the SG / simulink. The simulation results show that the reconfigurable technology based on FPGA has the following advantages, the strong load anti-interference, the fast overshoot time response, the utilization of resources can be effectively reduced, it provide a better a idea for hardware implementation of neural network in the field of power electronics.