Simulation and implementation of a 300 watt, cascade gama-LC resonant converter

In this paper a cascadeΓ - LC resonant power converter on base of a hybrid controller is introduced. To control the output voltage of converter, Artificial Neural Networks and PID classic controllers are combined. A multi-layer feed-forward artificial neural network is employed to achieve real-time control. The PID classic controller regulates the voltage on base of an Online PWM process while The ANN controller changes the switching frequency to maximize the voltage gain and consequently to increase the efficiency of converter. The resulted trained ANN was tested on an ideal simulated model of the converter for fine tuning. A prototype was designed and implemented on base of the proposed controller and was tested experimentally. Comparison between simulation and experimental results verified the proposed hybrid controller.