|
|||
Online Trained
Simulation and DSP Implementation of
Dynamic Back Abstract: A real time implementation aspect of backpropagation learning framework for training of a feed-forward neural network is proposed. The real time prototype of Back-Propagation based Dynamic Neuro Controller (BDNC) using DSP (TMS320LF2407A) is developed for buck converter in order to eliminate the tuning of the PID controller parameters. The detailed simulation process through Simulink/Matlab is also given. It can be applied to any linear or non-linear system. Software control of BDNC provides flexible control of neural network parameters like activation functions and learning rate. Simulation results are in accordance with the practical results. Keywords: Neural Network, Digital signal processor, buck converter, Robust control.
@ARTICLE{P1110609001,
AUTHOR = {S.G.
Kadwane and Amit kumar and B.M. Karan and
T Ghose},
TITLE = {Online Trained Simulation And DSP Implementation Of Dynamic Back Propagation Neural Network For Buck Converter}, JOURNAL = {ICGST International Journal on Automatic Control and Systems Engineering, ACSE}, YEAR = {2006}, VOLUME = {06}, ISSUE = {I}, PAGES = {27--34} }
|
|||
|
