Title: Parameter optimisation of human-simulated intelligent controller for a cart-double pendulum system

Authors: Guiqiang Chen, Linjian Tang, Zushu Li

Addresses: Institute of Commanding Automation, Chongqing Communication College, Zhi Hui Zi Dong Hua Jiaoyan Shi, Chongqing Tong Xin Xue Yuan, Chongqing, 400035, China. ' Institute of Commanding Automation, Chongqing Communication College, Zhi Hui Zi Dong Hua Jiaoyan Shi, Chongqing Tong Xin Xue Yuan, Chongqing, 400035, China. ' Institute of Intelligent Automation, Chongqing Automation College, Chongqing University, Chongqing, 400044, China

Abstract: Using a hierarchical and multi-mode control structure, a human-simulated intelligent controller (HSIC) successfully realises the swing-up and stabilisation control for a cart-double pendulum system. In order to effectively optimise several parameters in control laws and advance the further application space of HSIC, a hybrid simplex-genetic algorithm (HSGA) is proposed in this paper. By analysing the procedure of swing-up and stabilisation control, the paper discusses how we determine the solution space of parameters and select the fitness function suited to this optimisation problem. The two different strategies, which are HSGA strategy and manual adjustment method, are performed the simulation experiments, respectively. The simulation results show that it has the better response of HSIC obtained by the proposed HSGA strategy. So it is proved that the proposed HSGA is effective to implement parameter optimisation of HSIC.

Keywords: parameter optimisation; human-simulated intelligent control; HSIC; genetic algorithms; GAs; orthogonal design; simplex method; pendulum systems; stabilisation control; swing-up control; simulation.

DOI: 10.1504/IJMIC.2010.034571

International Journal of Modelling, Identification and Control, 2010 Vol.10 No.3/4, pp.194 - 201

Published online: 10 Aug 2010 *

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