Nonlinear system identification for predictive control using continuous time recurrent neural networks and automatic differentiation

Al-Seyab, Rihab ORCID: 0000-0001-6384-193X and Cao, Y (2008) Nonlinear system identification for predictive control using continuous time recurrent neural networks and automatic differentiation. Nonlinear system identification for predictive control using continuous time recurrent neural networks and automatic differentiation, 18 (6). pp. 568-581.

Preview

Text (Peer Reviewed Version) Al Sayeb, R.K. (2006) Nonlinear system identification for predicative control using continuous time recurent neural networks and automatic differentiation.pdf - Accepted Version Available under License Creative Commons Attribution Non-commercial No Derivatives 4.0. Download (1MB) | Preview

Abstract

In this paper, a continuous time recurrent neural network (CTRNN) is developed to be used in nonlinear model predictive control (NMPC) context. The neural network represented in a general nonlinear state-space form is used to predict the future dynamic behavior of the nonlinear process in real time. An efficient training algorithm for the proposed network is developed using automatic differentiation (AD) techniques. By automatically generating Taylor coefficients, the algorithm not only solves the differentiation equations of the network but also produces the sensitivity for the training problem. The same approach is also used to solve the online optimization problem in the predictive controller. The proposed neural network and the nonlinear predictive controller were tested on an evaporation case study. A good model fitting for the nonlinear plant is obtained using the new method. A comparison with other approaches shows that the new algorithm can considerably reduce network training time and improve solution accuracy. The CTRNN trained is used as an internal model in a predictive controller and results in good performance under different operating conditions.

Item Type:	Article
Article Type:	Article
Uncontrolled Keywords:	Nonlinear system; System identification; Predictive control; Recurrent neural network; Automatic differentiation
Subjects:	T Technology > T Technology (General) > T55 Industrial Engineering. Management engineering
Divisions:	Schools and Research Institutes > School of Education and Science
Research Priority Areas:	Applied Business & Technology
Depositing User:	Rihab Al Seyab
Date Deposited:	31 Oct 2019 16:06
Last Modified:	01 Sep 2023 12:00
URI:	https://eprints.glos.ac.uk/id/eprint/7474

University Staff: Request a correction | Repository Editors: Update this record