PS-FB-ZVZCS PWM Converter Modeling and Simulation Based on the PID Neural Network

Abstract

In order to overcome the disadvantage of the traditional control methods and general neural network control methods, the above two control methods which used to be applied to the PS-FB-ZVZCS-PWM(Phase-shifted Full-bridge Zero-voltage Zero-current-switching Pulse-Width Modulation, PS-FB-ZVZCS-PWM)converters modeling has been replaced by the PID(Proportional-Integral-Derivative , PID)neural network control. The first PID neural network subnet was used as the outer voltage loop control and the second PID neural network subnet was used as the inner current loop control. The output of the first PID neural network subnet was used as the reference input of the second PID neural network subnet. By the tight integration of two neural network subnets, a dual loop PID neural network control system was got. The result of the simulation which was got by MATLAB software showed the use of PID neural network as a regulator of the double close loop model was not only to achieve the nice control characteristics which are no overshoot, no static error, fast response, short transition time, good tracking performance, but also man-made regulation time was significantly reduced.