Motion Control For A Car-like Robot: Potential Field And Multi-agent Approaches

| This paper is concerned with the problem of controlling the motion of a mobile robot moving in a time-varying environment. The study considers a car-like vehicle moving in a roadway-like environment. We present two approaches which enable the vehicle to execute a given nominal motion plan and to adjust it in order to avoid static and dynamic obstacles. The rst approach is based on classical potential eld techniques. The second one relies upon a multi-agent approach. We are intersted in the problem of motion control of a car-like vehicle. To deal with this problem, we use a classical two layered system composed of a \trajectory planner" and a \motion controller". The purpose of the trajec-tory planner is to generate a nominal plan made up of a geometric path and a velocity proole along this path. The motion controller enables the autonomous vehicle to execute this nominal plan in a reactive way and to optimize its behaviour, i.e. to avoid moving obstacles while respecting the highway code and the kinematic constraints of the vehicle. In this paper, we present two approaches to implement the motion controller. In the rst scenario, developed for a partially known environment, the vehicle has a model of the environment (a road map) moreover a perception system provides it with information about its neighborhood. In this case, the motion controller uses a numerical potential eld approach ?] ?]. In the second scenario, developed for an unknown environment, the model of the local environment is incrementally built using local information obtained by the perception system of the vehicle. We use an adaptation of a multi-agent approach, the coordination patterns approach ?]. In order to solve the problem of motion control of a vehicle , we have developed a classical hierarchical system composed of a trajectory planner and a motion controller ?]. The trajectory planner provides the motion controller by a nominal motion plan made up of a geometric path and a velocity proole along this path ?, ?]. Let P be the nominal motion plan provided by the tra-jectory planner, M the vehicle considered and W the actual environment. P is made up of a geometric path |i.e. a curve of W| and a velocity proole for M. The motion controller of M has the task to generate the appropriate motion commands which are required to execute P while constantly adapting the behaviour of M to the …