International Conference on Control, Automation and Systems 2007 Oct. 17-20, 2007 in COEX, Seoul, Korea 1. INTRODUCTION Collaborative wheeled mobile robots can be modeled as an underactuated system having nonholonomic constraints imposed on their kinematics. Due to these constraints, the collaborative robots are not stabilizable around a point by continuous, time invariant feedback. Therefore, linear control is ineffective and innovative design techniques are required. One such possible design technique is feedback linearization [1-2]. Feedback linearization differs from Jacobian linearization, because feedback linearization is achieved using exact state transformations and feedback, rather than by linear approximations [3]. The various approaches used for formation control of collaborative robots can be roughly divided into three categories: leader-follower formation, behavior based and virtual structure formation [4-5]. In this paper, the robots area assumed to have communication capabilities only. So the leader-follower formation is used, in which the follower robots maintain a desired distance and a desired bearing angle relative to the leader robot [6]. A detailed analysis of control properties including controllability and stability were presented in [7]. Based on the control analysis, different feedback control strategies for the leader robot were developed. A comparative assessment of various feedback control strategies can be found in [8]. However, most of these feedback control strategies are efficient in trajectory tracking but do not address the issue of posture stabilization. In this paper, feedback linearized control strategies for both the leader and follower robots are presented. The main objective of the feedback linearized control strategies is to address the issue of posture stabilization for collaborative robots. A framework for collaborative robots is developed based on their kinematics. The framework is modeled using MATLAB/Simulink and simulated using the feedback linearized control strategies. The remainder of this paper is structured as follows. The kinematic model is presented in section 2. The development framework is discussed in section 3. Simulation results and conclusions are presented in section 4. Finally, summary and future work is discussed in section 5. 2. KINEMATIC MODELING OF LEADER FOLLOWER FORMATION A collaborative robot system can be described by its state which is a composition of the individual robots. The state of each robot varies as a function of its own state as well as the information sent by the other robots. Using the leader-follower formation for collaborative robots, the position of the leader robot can modeled using the vector p l = [x l , y l , θ l ] T . Let v l and ω l denote the leader robot’s linear and angular velocities, respectively. The kinematic equations for the leader robot can be expressed as Feedback Linearized Strategies for Collaborative Nonholonomic Robots Salman Ahmed 1 , Mohd N. Karsiti 2 and Ghulam M. Hassan 3 1,2 Department of Electrical and Electronic Engineering, Universiti Teknologi PETRONAS, Malaysia (Tel : +60-5-3687877; E-mail (s): salhmed@gmail.com, nohka@petronas.com.my) 3 Department of Computer Systems Engineering, NWFP University of Engineering and Technology, Pakistan (Tel : +92-91-9216590; E-mail: gmjally@nwfpuet.edu.pk) Abstract: Collaborative wheeled mobile robots are not stable around a point by continuous time-invariant feedback. Therefore, linear control is ineffective and innovative design techniques such as feedback linearization are required. This paper presents feedback linearized control strategies for collaborative nonholonomic robots using leader-follower formation. A framework for collaborative robots is developed based on their kinematics. The development framework relies on robots having communication capabilities instead of visual capabilities. The collaborative robot system is modeled using Simulink. From the simulation results, the full state linearized via dynamic feedback strategy for the leader robot globally stabilizes the system. Furthermore, the full state linearized via dynamic feedback strategy achieves postures stabilization for the leader-follower formation. For the follower robots, the input-output via static feedback linearized control strategies minimize the error between the desired and actual formation. Furthermore, the input-output linearized control strategies allow dynamical change of the formation at run-time. Thus, for a given feasible trajectory, the full state feedback linearized strategy for the leader robot and input-output feedback linearized strategies for the follower robots are found to be more efficient in stabilizing the system. Keywords: Feedback linearization, nonholonomic, leader-follower formation. 1551 978-89-950038-6-2-98560/07/$15 ICROS