This study proposes a measuring method to measure the altitude and straightness of hydraulic support groups based on point clouds to tackle the shortcomings in the current measuring methods for measuring the attitude and straightness of hydraulic support groups. First, we build an experimental platform and use three RGB-D cameras to obtain the point cloud of the hydraulic supports from different angles. Then, we calibrated the depth camera and determined the conversion relationship between the camera coordinate system and the world coordinate system. After point clouds preprocessing, we use the method of SAC-IA combined with ICP to conduct point clouds registration, to obtain the complete point cloud of the hydraulic support groups. We analyzed the geometric attitude of a single hydraulic support point cloud and selected characteristic points on each hydraulic support to solve the straightness. Finally, experimental verification is conducted. We did a comparative experiment between SAC-IA combined with the ICP method and ICP method, which proved that the former can effectively reduce the matching error; moreover, we conducted the hydraulic support groups attitude and straightness measurement experiment and compared it with the measured value of the sensors. We found that the errors are all within the acceptable range. It proves the feasibility of adopting point clouds to measure the attitude and straightness of the hydraulic support groups, which is important to realize the transparency and unmanned measuring of the fully mechanized mining face.

这项研究提出了一种基于点云的液压支撑群的高度和直线度的测量方法，以解决当前测量液压支撑群的姿态和直线度的测量方法的缺点。首先，我们建立一个实验平台，并使用三个RGB-D摄像机从不同角度获得液压支架的点云。然后，我们对深度相机进行了校准，并确定了相机坐标系和世界坐标系之间的转换关系。对点云进行预处理后，采用SAC-IA与ICP相结合的方法进行点云配准，以获得液压支撑组的完整点云。我们分析了单个液压支撑点云的几何姿态，并在每个液压支撑上选择了特征点以解决直线度问题。最后，进行了实验验证。通过SAC-IA结合ICP方法和ICP方法的比较实验，证明前者可以有效降低匹配误差。此外，我们进行了液压支撑组的姿态和直线度测量实验，并将其与传感器的测量值进行了比较。我们发现误差都在可接受的范围内。证明了采用点云测量液压支架群的姿态和直线度的可行性，对于实现综采工作面的透明性和无人值守测量具有重要意义。