A LLC Resonant Converter with Neural Network Controller for DC to DC Energy Conversion in Telecom Application

This paper presents development of a performance analysis of neural network controllers to over-come the appearance of nonlinearities and uncertainties in the LLC resonant converter. This paper addresses a LLC resonant converter circuits are built for low output voltages (24 V) and such circuits can be useful for Telecom and server application. For obtaining the DC voltage transfer function Fundamental Harmonic Approximation (FHA) method has been adopted. Important issues of this converter are physical size, high conversion ratio, efficiency, and startup. The proposed topology has energy conversion efficiency of 92%, when operated at fully loaded condition. A neural network controller based LLC isolated DC–DC converters are proposed to regulate over wide range of loads. Switching losses are minimized by zero voltage switching by the use of resonant inductor and capacitor. This converter has advantages like low switching loss, less EMI, and less switching stresses.