约束耦合影响下基于NMPC的差动机器人主动调速路径跟踪

NMPC-based active speed regulating path tracking for differential robots under the influence of constraints coupling

  • 摘要: 差动机器人相比其他类型的机器人展现出更高的灵活性和可扩展性,针对在差动机器人的路径跟踪控制中,差动机器人约束耦合导致路径跟踪控制精确性不足问题,本文以非线性模型预测控制(Nonlinear model predictive control, NMPC)为基础,通过分析NMPC中差动机器人纵向行驶速度与参考路径坐标点之间的耦合关系,提出了一种约束耦合影响下基于NMPC的差动机器人主动调速路径跟踪,实现了对差动机器人速度的主动动态调整。为了验证提出的控制方法,进行了Simulink仿真和实验验证。结果表明,该运动控制系统提升了差动机器人在路径跟踪控制上的精度,其中仿真位移误差的绝对值不超过 0.0723 m,航向误差的绝对值不超过0.0964 rad,相比恒速路径跟踪控制系统,能够将位移误差最大幅值减小99.22%,航向误差最大幅值减小93.32%。同时相比现有的主动调速路径跟踪控制系统,能够将位移误差最大幅值和航向误差最大幅值减小87.54%和29.69%。

     

    Abstract: Differential robots exhibit superior flexibility and scalability compared to other types of robots. Addressing the issue of reduced precision in path tracking control due to constraint coupling in differential robots, this study employs Nonlinear Model Predictive Control (NMPC) as the foundation. By analyzing the coupling relationship between the longitudinal velocity of the differential robot and the reference path coordinates within NMPC, we propose an NMPC-based active speed regulating path tracking for differential robots under the influence of constraints coupling, achieving proactive dynamic adjustment of the differential robot's velocity. To validate the proposed control method, Simulink simulations were conducted. The results demonstrate that the motion control system enhances the precision of differential robots in path tracking control, with the absolute value of simulated displacement error not exceeding 0.0723 m and the absolute value of heading error not exceeding 0.0964 rad. Compared to the constant speed path tracking control system, the proposed system can reduce the maximum amplitude of displacement error by 99.22% and the maximum amplitude of heading error by 93.32%. Meanwhile, compared with the existing active speed regulating path tracking control system, it is able to reduce the maximum magnitude of displacement error and the maximum magnitude of heading error by 87.54% and 29.69%.

     

/

返回文章
返回