Neural network-based optimal control of a DC motor positioning system
Maksym Khomenko and Volodymyr Voytenko
Industrial
Electronics Department,
Chernihiv, 14027,
E-mail: m.khomenko@inel.stu.cn.ua
E-mail: vvp@inel.stu.cn.ua
Yuriy Vagapov*
Department of
Engineering and Applied Science,
Plas Coch,
E-mail: y.vagapov@glyndwr.ac.uk
*Corresponding
author
Abstract: This article describes an optimal control algorithm
for a DC motor drive operating as a positioning system. The control algorithm is
based on combination of artificial neural network and state space method with
variable gain. The positioning system operating under proposed algorithm has
demonstrated a transient process close to optimal without overshoot. It has
been also shown that the control algorithm is robust to the change of the DC
motor parameters and the supply voltage disturbances. An experimental setup
based on TMS320F243 evaluation module has been designed and built in order to
prove the simulation results. The algorithm has been verified by simulation
using MATLAB/Simulink and proved by a number of practical experiments.
Keywords: DC electric drives; positioning systems; artificial
neural network; ANN; digital control.
Reference to this paper should be made as follows:
Khomenko, M., Voytenko, V. and Vagapov,
Y. (2013) ‘Neural network-based optimal control of a DC motor positioning
system’, Int. J.
Automation and Control, Vol. 7, Nos. 1/2,
pp.83–104.