摘要
针对当前仓储物流机器人自动定位技术视觉信息反馈延迟,无法确定深度信息,导致定位误差较大、延迟时间较长的问题。提出了基于视觉伺服的仓储物流机器人自动定位方法。通过几何地图描述全局环境模型,确定机器人结构,分析视觉伺服定位原理,根据雅可比矩阵,标定摄像机位置,利用Parzen窗方法预测像素空间密度,提取主成分,建立误差模型,实现仓储物流机器人自动定位。仿真实验证明,所提方法的定位误差较小,能够有效降低定位延迟时间,满足信息反馈实时性要求。
In view of the delay in the visual information feedback of the current automatic positioning technology of warehousing logistics robots,the depth information cannot be determined,which leads to the problems of large positioning errors and long delay times.An automatic positioning method of warehouse logistics robot based on visual servo is proposed.It describes the global environment model through geometric maps,determines the robot structure,analyzes the principle of visual servo positioning,calibrates the camera position according to the Jacobian matrix,uses the Parzen window method to predict the pixel space density,extracts the principal components,establishes the error model,and realizes the automatic positioning of the warehouse logistics robot.Simulation experiments prove that the positioning error of the proposed method is small,which can effectively reduce the positioning delay time and meet the real-time requirements of information feedback.
作者
宋哲
SONG Zhe(Avic Xi'an Flight Automatic Control Research Institute,Xi'an 710065 China)
出处
《自动化技术与应用》
2021年第8期70-74,共5页
Techniques of Automation and Applications
关键词
视觉伺服
仓储物流机器人
自动定位技术
雅可比矩阵
误差模型
visual servo
warehousing logistics robot
automatic positioning technology
Jacobian Matrix
error model
