Neural Network Based Sliding Mode Control for DC-DC Converters

Abstract:

Article Preview

DC-DC converters have some inherent characteristics such as high nonlinearity and time-variation, which often result in some difficulties in designing control schemes. RBF neural network sliding-mode control method is applied to PWM-based DC-DC converters in this paper. As the control input is duty cycle, the control inputs are in the scopes between 0 and 1. A general RBF neural network can not be suitable to control the PWM-based DC-DC converters, because the output layer of such network uses a linear function, and that the outputs of the network are between -∞ and +∞. Sigmoidal function is used instead of the output layer function in this paper to make the outputs are between 0 and 1. This sliding mode control method based on neural network can not only control the scope of the network output, but also eliminate the system chattering. Simulation experiments verify that this method can control the DC-DC converters well.

Info:

Periodical:

Advanced Materials Research (Volumes 211-212)

Edited by:

Ran Chen

Pages:

395-399

DOI:

10.4028/www.scientific.net/AMR.211-212.395

Citation:

L. P. Fan and Y. Z. Yu, "Neural Network Based Sliding Mode Control for DC-DC Converters", Advanced Materials Research, Vols. 211-212, pp. 395-399, 2011

Online since:

February 2011

Export:

Price:

$35.00

In order to see related information, you need to Login.

In order to see related information, you need to Login.