Design, Simulation and implementation of an adaptive controller on base of artificial neural networks for a resonant DC-DC converter

A new method of control for high-voltage Full Bridge Series-Parallel Resonant (FBSPR) DC-DC converter with capacitive output filter, using Artificial Neural Networks (ANN) is proposed in this paper. The output voltage regulation obtained via high switching frequency and Soft switching operation (ZCS and ZVS technologies) to decrease the losses and optimize the efficiency of converter. In the following sections, The small signal model of converter is used to simulate the dynamic behavior of real converter using Matlab software. It was also used to obtain ideal control signals which are the desired ANN inputs and outputs and were saved as a training data set. The data set is then used to train the ANN to mimic the behavior of the ideal controller. In fact the ANN controller is trained according to the small signal model of converter and the ideal operating points. To compare the performances of simulated and practical ANN controller, a prototype is designed and implemented and is tested for step changes in both output load and reference voltage. Comparison between experimental and simulations show a very good agreement and the reliability of ANN based controllers.