Title: Optimised data-driven terminal iterative learning control based on neural network for distributed parameter systems

Authors: Xisheng Dai; Lanlan Liu; Zhenping Deng

Addresses: The School of Electrical and Information Engineering, Guangxi University of Science and Technology, Liuzhou 545006, China ' The School of Electrical and Information Engineering, Guangxi University of Science and Technology, Liuzhou 545006, China ' The School of Information Science and Engineering, Chongqing Jiaotong University, Chongqing 400074, China; Chongqing Keyuan Energy Technology Development, Ltd., Chongqing 401147, China

Abstract: In this paper, a data-driven iterative learning control with neural network-based optimisation method for distributed parameter systems is presented to solve a class of problems caused by the imprecise mathematical model. The forward difference format is used to establish a linear relationship between input and output data, which is the only information available. However, this also leads to an unknown parameter matrix of the system. To overcome this problem, the radial basis function neural network is used to form a mapping relation from the desired output to the desired input, and the iterative learning algorithm of neural network weight is obtained by optimising the system performance indexes. Then, a detailed theoretical analysis based on composite energy function is given. Moreover, unlike traditional iterative learning control task tracking the whole trajectory, tracking time terminal is taken into account in this paper. Finally, simulation results show the feasibility of the theory.

Keywords: data-driven control; iterative learning control; ILC; distributed parameter systems; DPSs; neural network; convergence.

DOI: 10.1504/IJAAC.2021.116422

International Journal of Automation and Control, 2021 Vol.15 No.4/5, pp.463 - 481

Received: 11 May 2019
Accepted: 13 Nov 2019
Published online: 23 Jul 2021 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article