摘要
Finding an optimal trajectory from an initial point to a final point through closely packed obstacles, and controlling a Hilare robot through this trajectory, are challenging tasks. To serve this purpose, path planners and trajectory-tracking controllers are usually included in a control loop. This paper highlights the implementation of a trajectory-tracking controller on a stepper motor-driven Hilare robot, with a trajectory that is described as a set of waypoints. The controller was designed to handle discrete waypoints with directional discontinuity and to consider different constraints on the actuator velocity. The control parameters were tuned with the help of multi-objective particle swarm optimization to minimize the average cross-track error and average linear velocity error of the mobile robot when tracking a predefined trajectory. Experiments were conducted to control the mobile robot from a start position to a destination position along a trajectory described by the waypoints. Experimental results for tracking the trajectory generated by a path planner and the trajectory specified by a user are also demonstrated. Experiments conducted on the mobile robot validate the effectiveness of the proposed strategy for tracking different types of trajectories.
在障碍物密集的环境中,找到一条从初始位置到目标位置的最优轨迹,并控制一台Hilare机器人沿着该轨迹行驶仍是一项具有挑战性的任务。为了完成这个任务,控制环中通常需要加入路径规划器以及轨迹跟踪控制器。本文的目的是在一台由步进电机驱动的Hilare机器人上实现轨迹跟踪控制的任务。其中,轨迹由航点集合表示。在设计过程中,控制器需要考虑处理方向连续的离散航点,并且需要考虑不同的执行器速度约束。本文利用多目标粒子群优化(multi-objective particle swarm optimization, MOPSO)的方法来调整控制器的参数。MOPSO通过最小化移动机器人在追踪预定义轨迹时的平均航迹误差以及平均线速度误差来得到最优的控制器参数。实验中,移动机器人被控制从起始点沿着一条由航点表示的轨迹行驶到达目标点。实验同样给出对路径规划器生成的轨迹,以及自定义轨迹的跟踪结果。基于移动机器人的实验结果验证了本文方法对不同形式轨迹跟踪的有效性。
