基于深度视觉传感的巡检机器人行进轨迹跟踪控制  

Tracking Control of Inspection Robot Based on Depth Vision Sensing

作　　者：郭建仙 GUO Jianxian(CHN Energy Shenfu(Shishi)Power Generation Co.,Ltd.,Quanzhou 362700,China)

机构地区：[1]国能神福(石狮)发电有限公司,福建泉州362700

出　　处：《机械与电子》2023年第9期53-56,62,共5页Machinery & Electronics

摘　　要：巡检机器人面对不同路况时反应速度的时间差使其行进轨迹跟踪控制容易产生误差。为了提高巡检机器人复杂操作环境中的行进轨迹跟踪控制效果,提出基于深度视觉的巡检机器人行进轨迹跟踪控制方法。根据巡检机器人移动位姿和轴向运动轨迹产生的角速度和线速度,建立巡检机器人运动学方程;利用Kinect深度视觉传感器创建贴合实际环境的三维地图;利用滑模控制器约束巡检机器人运动学方程,实现巡检机器人行进轨迹自动化跟踪控制。实验结果表明,所提方法在障碍物环境和非障碍物环境的跟踪控制效果均较好。The time difference of the reaction speed of the inspection robot when facing different road conditions makes the tracking control of its travel path prone to errors.In order to improve the tracking control effect of the inspection robot in the complex operating environment,a method based on depth vision for the tracking control of the inspection robot is proposed.Firstly,the kinematics equations of the inspection robot are established according to the angular velocity and linear velocity generated by the moving posture and axial motion trajectory of the inspection robot;Kinect depth vision sensor is used to create a 3D map that fits the actual environment;the sliding mode controller is used to constrain the kinematic equation of the inspection robot,and the automatic tracking control of the inspection robot’s travel path is realized.The experimental results show that the proposed method is effective in both obstacle and non obstacle environments.

关 键 词：巡检机器人 运动学方程 Kinect深度视觉传感器 行进轨迹 滑模控制器 

分 类 号：TP242[自动化与计算机技术—检测技术与自动化装置]