世界坐标系下非完整轮式移动机器人转弯轨迹的时域非微分描述方程

为了解决机器人路径规划模块在为非完整轮式移动机器人规划路径时,若要实现规划轨迹平滑且完全满足机器人的运动学约束,则需要精确的机器人工作环境的地图信息,从而大幅度地增加了路径规划算法的计算量的问题,首先利用较为粗略的环境地图信息规划路径,再将机器人转弯部分的路径进行函数插值,使整个路径规划满足机器人运动学约束且计算量增加较少.现有的机器人轨迹描述方程不能满足对机器人转弯轨迹进行精确插值的需要,因此提出分别利用曲线拟合法和泰勒插值法来构建非完整轮式移动机器人转弯轨迹的时域非微分描述方程.实验结果显示,所述机器人轨迹非微分时域描述方程可以用于对机器人路径规划的插值,降低了算法的计算量.

Non-differential Equations for Describing Trajectory of Nonholonomic Wheeled Mobile Robot in World Coordinates and Its Application

When a robot path planning module plans a path for nonholonomic wheeled robots, if we want to make the robot trajectory smooth and completely meet the robot kinematics constraints, we need the accurate map information of robot working environment which significantly increases the calculation amount of path planning algorithm. To solve the problem of the increased amout of calculation, a rough environment map information was adopted to plan a path in this paper. Then the interpolation function to represent the turning part was adopted. The solution ensured the whole path planning meet the robot kinematics constraints and increase the minimum amount of calculation. The existing robot trajectory equations cannot satisfy the need of robot turning trajectory for precise interpolation. So a curve fitting method and Taylor interpolation method were put forward to construct a nonholonomic wheeled mobile robot trajectory of turning time domain of differential equations. Experimental results show that the proposed methods can be used to interpolate the planning results of path and reduce the computational cost of the algorithm.