基于预测控制的全向移动机器人轨迹跟踪

Trajectory Tracking Control for Omnidirectional Mobile Robot Using Predictive Control

针对全向移动机器人的轨迹跟踪控制问题,在分析与建立全向移动机器人运动学与动力学模型的基础上,研究了基于模型预测控制的移动机器人轨迹跟踪方法;通过对误差模型的线性化描述,并将目标函数以二次项部分与线性部分构成,在满足控制约束的条件下,将采样周期内最小化目标函数的优化问题转换为二次规划问题的求解;当取预测时域N=5时,机器人在x和y方向的误差范围分别为±5.4%和±4.7%,并在轨迹曲线中曲率半径较小处出现较为明显的抖动,最后对误差产生原因进行了分析与总结,结果表明该方法对全向移动机器人的轨迹跟踪控制是有效可行的。

Aiming at solving the motion control problem of omnidirectional mobile robot,after analyzing and building the kinematics and dynamics models of the robot,the model predictive control method is applied to the trajectory tracking problem.By linearize the error models,,the objective function can be made up of quadratic part and linear part.Then the optimization problem of minimizing the objective function is formulated as a quadratic programming with control constrains satisfaction.While taking prediction horizon N=5,the error range along x direction and y direction were no more than 5.4% and 4.7% respectively,and the robot will suffer some dithering at small curvature radius on trajectory curve.Finally,based on summarizing the errors caused in this experiments,the results show the validity and feasibility for trajectory tracking of the omnidirectional mobile robot.