基于机器视觉的悬臂式掘进机机身位姿检测系统

针对矿井巷道自动掘进的需求,为矿用悬臂式掘进机构建了一套机身位姿实时检测系统。该系统以十字激光器与激光标靶为信息来源,通过对标靶上十字光线成像特征的分析,建立了掘进机机身位姿空间解算模型。该模型利用机身与十字激光面的空间关系,使用空间矩阵变换方法,得到机身相对于巷道的三轴倾角以及在巷道断面上的偏离位移,实现了掘进机机身位姿的自动实时检测。最终,在实验室条件下,搭建了机身位姿自动检测试验平台,模拟掘进机在巷道内的姿态,试验结果表明:测量范围在2~100 m时,系统的角度测量误差在0.5°范围内,位移的检测误差小于20 mm,能够满足巷道施工过程中掘进机机身位姿自动、精确、实时测量的要求。

Pose measurement system of boom-type roadheader based on machine vision

In order to meet the needs of roadway automatically drivage under mine,a new system,real-time posture and position detection scheme is proposed for the mine-used boom-type roadheader.By regarding cross lasers and laser targets as information sources,this system establishes a mathematical model to describe the machine position in space.Utilizing machine vision technology to extract image feature parameters,the authors realize the precise localization of reference points on targets.After setting up the machine body position calculation model,the system takes use of the space matrix transformation method and obtains the yaw,pitch,roll angles and horizontal,vertical deviations of the roadheader.Finally,the automatic machine pose detection in real time is fulfilled.On the basis of above theoretical analysis,the authors build up the position automatic detection experiment platform in laboratory conditions to simulate the pose changes of the machine inside the tunnel.The experimental results show that the measurement accuracy of angles is within 0.5 degree and the detection precision of displacement is less than 20 mm,which satisfies the requirements of roadheader automatic,precise and real-time positioning in the process of tunnel construction.