Authors: Anmin Zhu, Simon X. Yang
Addresses: College of Information Engineering, Shenzhen University, Shenzhen, Guangdong, China School of Engineering, University of Guelph, Guelph, Ontario, N1G 2W1, Canada. ' Advanced Robotics and Intelligent Systems (ARIS) Laboratory, School of Engineering, University of Guelph, Guelph, Ontario, N1G 2W1, Canada
Abstract: A novel tracking controller for real-time navigation of a non-holonomic mobile robot is presented. The proposed tracking controller is based on a backstepping model and a gated dipole model from a biological neural system. It is capable of generating smooth, bounded acceleration control signals to track a reference path. Unlike the conventional backstepping controllers, the proposed model can solve the velocity jump problem caused by a large initial tracking error. Unlike the existing tracking control methods that assume |perfect velocity tracking|, the proposed controller removes the unpractical assumption by considering the constraint of accelerations. In addition, the proposed controller can deal with complicated situations, e.g. the reference linear velocity or angular velocity is not constant. The effectiveness and efficiency of the approach are demonstrated by simulation studies.
Keywords: gated dipole model; neural dynamics; non-holonomic robots; mobile robots; tracking control; robot control; robot navigation; simulation; backstepping model; vehicle autonomous systems; autonomous vehicles.
International Journal of Vehicle Autonomous Systems, 2006 Vol.4 No.2/3/4, pp.156 - 172
Available online: 28 Jan 2007 *Full-text access for editors Access for subscribers Purchase this article Comment on this article