基于视觉图像的全向移动机器人轨迹跟踪控制研究

Research on omnidirectional mobile robot trajectory tracking control based on visual image

摘要:

机器人轨迹节点跟踪比较难，导致机器人实际轨迹偏离期望轨迹，所以设计基于视觉图像的全向移动机器人轨迹跟踪控制方法。以滑模变结构控制闭环为基础，求解运动学模型、动力学模型表达式，实现对机器人移动轨迹数学模型的构建。按照视觉图像划分标准，完成对全向移动机器人运动图像的分割，通过分离目标节点的方式提取运动学特征参量，完成机器人轨迹节点跟踪处理。设置前馈控制器与扰动观测器，根据运动学不等式条件计算误差向量指标的取值范围，根据该值对主机元件的控制作用能力进行调节，实现全向移动机器人轨迹跟踪控制。对比实验结果表明，所设计的方法应用后，全向移动机器人角速度曲线、线速度曲线与期望运动轨迹曲线之间的贴合程度均超过90%，满足全向移动机器人轨迹跟踪控制要求。

Abstract:

It is difficult to track the trajectory node of the robot, which leads to the deviation of the actual trajectory of the robot from the desired trajectory. Therefore, an omnidirectional mobile robot trajectory tracking control method based on visual image is designed. Based on the sliding mode variable structure control loop, the kinematic model and dynamic model expression are solved, and the mathematical model of robot moving trajectory is constructed. According to the visual image segmentation standard, the motion image of omnidirectional mobile robot is segmented, and the kinematic feature parameters are extracted by separating the target nodes to complete the tracking process of robot trajectory nodes. The feedforward controller and disturbance observer were set, and the value range of the error vector index was calculated according to the kinematic inequality condition. According to the value, the control ability of the host element was adjusted, and the omnidirectional mobile robot trajectory tracking control was realized. The experimental results show that after the application of the proposed method, the fitting degree between the angular velocity curve, the linear velocity curve and the desired trajectory curve of the omnidirectional mobile robot is more than 90%, which meets the trajectory tracking control requirements of the omnidirectional mobile robot.