单线激光雷达扫描的矿车装载量核算方法

面向露天矿山生产对于矿车装载量核算的需求，本文提出了一种利用单线激光雷达扫描的矿车装载量自动核算方法。该方法首先采用固定式单线激光雷达对目标车辆进行俯视扫描，基于车辆形态特征分割出载荷物顶部点云；然后经时态校正，并与预先构建的车辆信息数据库进行匹配，重建出载荷物的三维表面模型；最后通过对载荷物表面模型的切片积分实现方量核算。以齐大山铁矿运输为例，对多台矿车装载量进行核算测试。结果表明，本文方法可在车辆正常行驶的情况下实时获取载荷物的表面三维点云，平均建模误差为2.54 cm，装载方量核算误差为1.17%。

Mining vehicles loading estimation by single-pulse laser scanning

Facing the demand of loading volume estimation for vehicles in open pit mines, this paper proposes an automatic measuring method using single-pulse laser scanning. First, we use a fixed single pulse laser to scan the target vehicle in a bird view, and then segment the point cloud of the load surface using the vehicle morphological features. Following that, we reconstruct the load surface model by temporal correction and matching the pre-constructed vehicle database, and finally estimate the loading volume by slicing and integrating the surface model. We apply the proposed method to Qidashan iron ore open pit mine for load volume estimation. Experimental results show that the proposed method enables to obtain the surface 3D point cloud of the load in real time under the moving conditions of vehicles. Besides, the average error of 3D modeling is 2.54 cm, whereas the error of load volume estimation is 1.17%.