A neuro-fuzzy controller applying to a Cuk converter

A DC-DC power converter is difficult to control due to its nonlinearities and parameter uncertainties. To tackle this problem, a neuro-fuzzy controller is proposed. The controller utilizes the error voltage and the change of error voltage as inputs, and outputs the duty cycle of the PWM switch for controlling the power converter. Instead of relying on expert knowledge, some heuristic rules are derived using the membership functions of the fuzzy variables tuned by a neural network. After offline training, the neuro-fuzzy controller can be applied to regulate a Cuk power converter. Simulation results are to be given to demonstrate the performance of the controller