Sliding mode control of a DC/DC PWM converter with PFC implemented by neural networks

Abstract

An experimental neural controller implementing a variable structure control (VSC) algorithm is proposed for a power factor preregulator. VSC control laws yield fast response and a robust behavior against large parameters variations. A multilayer perceptron learns through backpropagation to approximate the desired adaptive control functions. The main advantage of the neural network implementation in comparison to the numerical implementation is that decreases complexity and cost of the controller, and increases the switching frequency. A simple analog electronic realization of this neural network using discrete operational amplifiers is proposed. This implementation possesses all good properties of sliding mode while avoiding the unnecessary discontinuities of the control input signals and thus eliminating chattering. Experimental results are summarized confirming the validity of the neural network approach.

General Design Issues of Sliding-Mode Controllers in DC–DC Converters

“…In 1996, Fossas and Biel [39] presented a new approach for the design of SM-controlled dc-dc converters for generating an ac signal. In 1997, Carrasco et al [40] proposed to incorporate neural networks with SM controllers for power-factor-correction applications.…”

Power Electronic Converters Modeling and Control

“…Articles like Venkataramanan et al (1985), Sira-Ramírez (1987, Malesani et al (1995), Spiazzi et al (1995), Carpita and Marchesoni (1996), , Carrasco et al 1997;Martínez-Salamero et al (1998), Guffon et al (1998), Sira-Ramírez (2003 and Tan et al (2005) are just few of the bibliographical references that the literature provides. There are also some notable textbooks like Sira-Ramírez and Silva-Ortigoza (2006) and Tan et al (2011) that provide useful theoretical and practical insights in developing variable-structure control structures for power electronic converters.…”

A Beta basis function neural network in CMOS subthreshold mode

“…A high density is of primary importance in analogue neural networks and analogue VLSI (Cichocki et al 1992, Carrasco et al 1997, Masmoudi et al 1999). …”