Sliding-Mode Control and Sonnar Based Bubble Rebound Obstacle Avoidance for a WMR Adrian Filipescu, Bogdan Dumitrascu, Adriana Filipescu, George Ciubucciu Department of Automation and Electrical Engineering “Dunarea de Jos” University of Galati, Romania Adrian.Filipescu@ugal.ro Eugenia Minca Department of Automation, Computer Science and Electrical Engineering “Valahia” University of Targoviste, Romania eugenia.minca@gmail.com Alina Voda Grenoble Image Parole Signal Automatique (GIPSA-lab), University Joseph Fourier Grenoble 1/CNRS, UMR 5216, B.P. 46, F-38402 St Martin d’Heres, France alina.voda@gipsa-lab.grenoble-inp.fr Abstract— In this paper an algorithm for trajectory-tracking and obstacle avoidance for wheeled mobile robots (WMR) is presented. The algorithm creates a trajectory composed of a global trajectory generated off-line and local obstacle avoidance trajectories that are created when an obstacle is detected by the sonar sensors. Only one discrete-time sliding-mode controller is required to track the resulting trajectory and it does not require separate controllers for following the intended trajectory and avoiding the obstacle. The local avoidance trajectories are generated using Quintic equations to generate a path for the robot and assigning calculating the velocity, acceleration, angular velocity and angular acceleration needed by the discrete-time sliding-mode controller. Keywords— sliding-mode, obstacle avoidance, WMR. I. INTRODUCTION WMRs have been used in a multitude of applications, especially for industrial applications. Some of these applications require the robots to track a set trajectory. Occasionally temporary obstacles, like a broken robot, or fallen debris, can appear and the robots will be unable to follow the desired trajectory leading to a full stop of the entire work. A solution to this problem is to include an obstacle avoidance module that will allow the robot to avoid the blocked part of the trajectory and to return to the desired trajectory. This solution will add a delay to the process but will prevent a full stop of the process. Sliding-mode control (SMC) is a robust approach for various applications and it can be used. The SMC methodology has been presented in [1]. Sliding-mode has been used in many control systems, such as: [2]-[16]. The discrete-time sliding-mode control was chosen to solve the trajectory-tracking problem because it promises great results. One of the simplest obstacle avoidance algorithm is “the bug”, proposed by in [20]. This algorithm requires the robot to circle the obstacle before determining the best point to continue towards its goal. Other obstacle avoidance algorithms use virtual potential fields, like in [18] to determine the path the robot has to take in order to avoid the obstacle. The algorithms consider that the robot is subjected to forces that pull the robot towards the goal and rejecting forces that pull the robot away from the obstacles. The robot’s trajectory results from the combination of these forces. A vector field histogram is used in [17] for obstacle avoidance, that uses a histogram to reduce the errors from the sensors. A bubble rebound algorithm is used in [19] to avoid obstacles. The algorithm uses a sensitivity bubble to determine if an obstacle is blocking the path. If an obstacle is detected the algorithm calculates the path that has the minimum density of obstacles and moves in that direction until the goal is visible or a new obstacle is detected. An algorithm that uses capable of tracking a desired trajectory and avoids obstacles that block the desired path by generating local trajectories and following them until the initial trajectory can be resumed is proposed in this paper. The global and local trajectories will be merged and form a new trajectory that can be tracked using only one discrete-time sliding-mode controller. The robot used to test the algorithm is the two driving wheels/two free wheels (2DW/2FW), PowerBot (Fig. 1). The WMR PowerBot is a high-payload differential-drive robot capable of moving at speeds up s m / 6 . 1 and carrying up to 100 kg. The obstacles are detected using the sonar sensors of the robot and the sensitivity bubble from [19]. The rest of the paper is organized as follows: discrete-time, sliding-mode control based on kinematic model of the WMR is presented in Section II; bubble rebound obstacle avoidance algorithm for the 2DW/2FW WMR PowerBot is presented in Section III; simulation results of obstacle avoidance, trajectory tracking control are presented in Section IV; some final remarks can be found in Section V. II. DISCRETE-TIME SLIDING-MODE CONTROL Consider the model of a wheeled mobile robot, presented in Fig. 2. The model takes into account the two diametrically opposed drive wheels of radius R, the distance between the 2015 19th International Conference on System Theory, Control and Computing (ICSTCC), October 14-16, Cheile Gradistei, Romania 978-1-4799-8481-7/15/$31.00 ©2015 IEEE 105