摘要
介绍了一种基于雷达的长壁采煤机械定位系统,该系统基于雷达测距传感器,用于确定采矿设备相对于矿山煤巷基础设施的位置,通过试验验证了使用雷达传感器进行定位的合理性;为了从单个雷达信号中估计两个关键的定位参数,即沿轨位置和跨轨位置,研究了几种概率数据处理技术,对于跨轨位置,传统的卡尔曼滤波方法足以实现可靠的估计,对于沿轨位置估计,必须通过跟踪算法来识别煤巷肋墙上的特定基础设施元素,我们在三维交互显示中探索了一种新的可视化分析方法以方便识别重要特征用于分类器算法,基于分类器的输出,使用已识别的元素作为位置路径点,可以提供一个稳定和准确的采矿设备定位估计。
A long wall coal mining machinery positioning system based on radar is introduced, the system is based on radar ranging sensor to determine the position of mining equipment relative to mine coal roadway infrastructure. The positioning rationality by using the radar sensor is verified by the experiment. In order to estimate two key positioning parameters from single radar signal, i.e.along track position and cross track position, several probabilistic data processing technologies are studied. For the cross track position, traditional Kalman filtering is enough to achieve reliable estimation. For the along track position estimation, specific infrastructure elements on the rib wall of coal roadway must be identified by tracking algorithm. A new visual analysis method in 3D interactive display is explored to facilitate the identification of the important features for the classifier algorithm. On the basic output of classifier, the identified elements are taken as the location path points, can provide a stable and accurate location estimation of mining equipment.
作者
安文利
李国强
孙炜歆
AN Wenli;LI Guoqiang;SUN Weixin(State-owned Baorixile Energy Co.,Ltd.,Hulunbuir 021000,China)
出处
《计算机测量与控制》
2023年第1期147-152,共6页
Computer Measurement &Control
关键词
定位
路标导航
机器学习
雷达
地下
长壁采矿
自动化
positioning
landmark navigation
machine learning
radar
underground
longwall mining
automation
