Abstract:Aiming at the problem of nonholonomically constrained motion planning of Ackerman robot, a Timed Elastic Band (TEB) algorithm was proposed by Christoph R?smann et al. to implement the local path planner of the robot. The TEB algorithm optimized the robot trajectory by modifying the initial trajectory generated by global path planning. When the rate of change of the acceleration was too large, the sudden change in the torque output by the robot's chassis motor would cause the robot to be shocked. The constraint of jerk could limit the rate of change of acceleration to a reasonable range. Based on the lack of jerk con-straint in the TEB method, in the trajectory optimization process, specific jerk constraint was constructed in this paper. The improved TEB algorithm was simulated on the Stage simulation platform and implemented on a real Ackerman robot. The simulation and experimental result showed that the obtained smooth tra-jectory met the requirement of kinematics, kinodynamics, and avoided obstacles. It can be seen that the improved TEB algorithm is suitable for Ackerman robots, and the effect of planned trajectory is good.