Title :
Design of blender IMC control system based on simple recurrent networks
Author :
Du, Yun ; Sun, Hui-qin ; Tian, Qiang ; Zhang, Su-Ying ; Wang, Chang
Author_Institution :
Hebei Univ. of Sci. & Technol., Shijiazhuang, China
Abstract :
Blender is the key tache in the production process of foodstuff and chemistry and so on. The material control value has a complicated non-linear relationship with the flux value. It is difficult to build the accurate mathematical model by traditional method. The internal model control (IMC) system based on the simple dynamic recurrent neural network (SRNN) is presented in this paper. It is very simple because of the recursion layer without the weight values in SRNN, and the regulated network parameters is very few. The control system based on SRNN is composed of the positive model, the inverse model and the filter. The simulation shows that the system has the better anti- disturbing capability. It can satisfy the rapidness and the veracity of control system. It can be used in real time.
Keywords :
blending; control system synthesis; production engineering computing; recurrent neural nets; antidisturbing capability; blender IMC control system; dynamic recurrent neural network; flux value; internal model control; inverse model; material control value; mathematical model; nonlinear relationship; production process; simple recurrent networks; Chemistry; Control systems; Convergence; Cybernetics; Feedforward neural networks; Feeds; Machine learning; Neural networks; Nonlinear dynamical systems; Recurrent neural networks; Blender; Dynamic recurrent neural network; Internal model control;
Conference_Titel :
Machine Learning and Cybernetics, 2009 International Conference on
Conference_Location :
Baoding
Print_ISBN :
978-1-4244-3702-3
Electronic_ISBN :
978-1-4244-3703-0
DOI :
10.1109/ICMLC.2009.5212450
