This paper describes a method to make a wheeled mobile robot follow a given path just like a driver would decelerate the speed of a car before a corner and accelerate it on the straight course of the road. Two kinds of velocity vectors are introduced to determine an optimal direction and speed of the robot in order to follow the path. The velocity defined as a function of curvature of the path is mainly utilized to change the speed of the robot along the path. The other velocity given as the tangent of a virtual potential function produces an attraction force toward the path. Path tracking control of the robot is formulated with current and future sets of these velocity vectors in a frame of the digital preview and predictive control. By this method, the robust tracking performance is achieved.
経路の未来情報を利用した車輪型自立移動ロボットの経路追従制御
12_4_630.pdf
(3.39MB)
