随着人工智能和无人驾驶等相关学科的快速发展,煤矿装备的智能化和无人化成为了新的趋势。智能设备的应用将大幅提高煤矿作业的生产力以及人员安全性。露天煤矿地形复杂,与城市环境相比无明显的几何特征,具有分段相似性,利用现有以激光...随着人工智能和无人驾驶等相关学科的快速发展,煤矿装备的智能化和无人化成为了新的趋势。智能设备的应用将大幅提高煤矿作业的生产力以及人员安全性。露天煤矿地形复杂,与城市环境相比无明显的几何特征,具有分段相似性,利用现有以激光雷达为主的同时定位与建图(Simultaneous Localization and Mapping,SLAM)方案在该环境下易出现定位漂移和建图误差较大等现象。针对上述问题,提出了一种基于激光雷达(Light Detection and Ranging,LiDAR)和惯导(Inertial Measurement Unit,IMU)紧耦合的SLAM算法,该算法使用LiDAR和IMU两种传感器作为数据输入,对数据进行预处理,前端利用迭代扩展卡尔曼滤波器将预处理后的LiDAR特征点与IMU数据相融合,并使用后向传播来矫正雷达运动畸变,后端利用雷达相对位姿因子将LiDAR帧间配准结果作为约束因子与回环因子共同完成全局因子图优化。利用开源数据集和露天煤矿实地数据集验证了算法的鲁棒性和精确性。试验结果表明在城市结构化环境中文中所提算法与当前激光SLAM算法精度保持一致,而针对长达两千多米的露天煤矿实地环境,所提算法较FAST-LIO2、LIO-SAM紧耦合算法在定位精度上分别提高了46.00%和23.15%,且具有更高的鲁棒性。展开更多
Coal bumps have long been a safety hazard in coal mines, and even after decades of research, the exact mechanics that cause coal bumps are still not well understood. Therefore, coal bumps are still difficult to predic...Coal bumps have long been a safety hazard in coal mines, and even after decades of research, the exact mechanics that cause coal bumps are still not well understood. Therefore, coal bumps are still difficult to predict and control. The LaModel program has a long history of being used to effectively analyze displacements and stresses in coal mines, and with the recent addition of energy release and local mine stiffness calculations, the LaModel program now has greatly increased capabilities for evaluating coal bump potential. This paper presents three recent case histories where coal stress, pillar safety factor, energy release rate and local mine stiffness calculations in LaModel were used to evaluate the pillar plan and cut sequencing that were associated with a number of bumps. The first case history is a longwall mine where a simple stress analysis was used to help determine the limiting depth for safely mining in bump-prone ground. The second case history is a room-and-pillar retreat mine where the LaModel analysis is used to help optimize the pillar extraction sequencing in order to minimize the frequent pillar line bumps. The third case history is the Crandall Canyon mine where an initial bump and then a massive pillar collapse/bump which killed 6 miners is extensively back-analyzed. In these case histories, the calculation tools in LaModel are ultimately shown to be very effective for analyzing various aspects of the bump problem, and in the conclusions, a number of critical insights into the practical calculation of mine failure and stability developed as a result of this research are presented.展开更多
煤矿智能化是煤炭行业高质量发展的技术支撑,关键岗位的机器人替代是实现煤炭少人化、无人化的高效开采的发展趋势。即时定位与地图构建(Simultaneous Localization and Mapping,SLAM)是煤矿机器人自主移动与导航的关键技术之一。煤矿...煤矿智能化是煤炭行业高质量发展的技术支撑,关键岗位的机器人替代是实现煤炭少人化、无人化的高效开采的发展趋势。即时定位与地图构建(Simultaneous Localization and Mapping,SLAM)是煤矿机器人自主移动与导航的关键技术之一。煤矿井下为典型非结构化环境,空间狭长局促,结构复杂多变,照明情况不均匀,对煤矿井下SLAM提出了严峻挑战。总结了煤矿井下地图构建研究现状,针对LeGO-LOAM算法的回环检测仍存在的不足,利用SegMatch算法改进LeGO-LOAM的回环检测模块,且使用ICP算法进行全局图优化,提出了一种融合LeGO-LOAM和SegMatch的改进算法,阐述了该算法的原理和实现步骤;开展了煤矿井下模拟场景试验,对比分析改进前后SLAM算法的建图效果以及精度,试验结果表明改进算法构建的地图回环效果更好,估计轨迹更平滑、精确;结合导航需求研究了二维占据栅格地图的构建方法,试验验证了该方法所构建的栅格地图精度,结果表明有效滤除动态障碍物等离群噪点后的栅格地图具有0.01 m的建图精度,且所需存储空间较点云地图降低了3个数量级。研究成果有助于煤矿井下非结构环境下SLAM和煤矿机器人实时定位和自主移动。展开更多
文摘随着人工智能和无人驾驶等相关学科的快速发展,煤矿装备的智能化和无人化成为了新的趋势。智能设备的应用将大幅提高煤矿作业的生产力以及人员安全性。露天煤矿地形复杂,与城市环境相比无明显的几何特征,具有分段相似性,利用现有以激光雷达为主的同时定位与建图(Simultaneous Localization and Mapping,SLAM)方案在该环境下易出现定位漂移和建图误差较大等现象。针对上述问题,提出了一种基于激光雷达(Light Detection and Ranging,LiDAR)和惯导(Inertial Measurement Unit,IMU)紧耦合的SLAM算法,该算法使用LiDAR和IMU两种传感器作为数据输入,对数据进行预处理,前端利用迭代扩展卡尔曼滤波器将预处理后的LiDAR特征点与IMU数据相融合,并使用后向传播来矫正雷达运动畸变,后端利用雷达相对位姿因子将LiDAR帧间配准结果作为约束因子与回环因子共同完成全局因子图优化。利用开源数据集和露天煤矿实地数据集验证了算法的鲁棒性和精确性。试验结果表明在城市结构化环境中文中所提算法与当前激光SLAM算法精度保持一致,而针对长达两千多米的露天煤矿实地环境,所提算法较FAST-LIO2、LIO-SAM紧耦合算法在定位精度上分别提高了46.00%和23.15%,且具有更高的鲁棒性。
基金Projects(52274128, 51904167, 52174159) supported by the National Natural Science Foundation of ChinaProject supported by the Taishan Scholars Project Special Fund of Shandong Province,China+1 种基金Project(KCF2204) supported by the Open Fund for the Henan Key Laboratory for Green and Efficient Mining&Comprehensive Utilization of Mineral Resources,ChinaProject(22KF01) supported by the Open Fund for the State Key Laboratory of Mining Response and Disaster Prevention and Control in Deep Coal Mines,China。
文摘Coal bumps have long been a safety hazard in coal mines, and even after decades of research, the exact mechanics that cause coal bumps are still not well understood. Therefore, coal bumps are still difficult to predict and control. The LaModel program has a long history of being used to effectively analyze displacements and stresses in coal mines, and with the recent addition of energy release and local mine stiffness calculations, the LaModel program now has greatly increased capabilities for evaluating coal bump potential. This paper presents three recent case histories where coal stress, pillar safety factor, energy release rate and local mine stiffness calculations in LaModel were used to evaluate the pillar plan and cut sequencing that were associated with a number of bumps. The first case history is a longwall mine where a simple stress analysis was used to help determine the limiting depth for safely mining in bump-prone ground. The second case history is a room-and-pillar retreat mine where the LaModel analysis is used to help optimize the pillar extraction sequencing in order to minimize the frequent pillar line bumps. The third case history is the Crandall Canyon mine where an initial bump and then a massive pillar collapse/bump which killed 6 miners is extensively back-analyzed. In these case histories, the calculation tools in LaModel are ultimately shown to be very effective for analyzing various aspects of the bump problem, and in the conclusions, a number of critical insights into the practical calculation of mine failure and stability developed as a result of this research are presented.
文摘煤矿智能化是煤炭行业高质量发展的技术支撑,关键岗位的机器人替代是实现煤炭少人化、无人化的高效开采的发展趋势。即时定位与地图构建(Simultaneous Localization and Mapping,SLAM)是煤矿机器人自主移动与导航的关键技术之一。煤矿井下为典型非结构化环境,空间狭长局促,结构复杂多变,照明情况不均匀,对煤矿井下SLAM提出了严峻挑战。总结了煤矿井下地图构建研究现状,针对LeGO-LOAM算法的回环检测仍存在的不足,利用SegMatch算法改进LeGO-LOAM的回环检测模块,且使用ICP算法进行全局图优化,提出了一种融合LeGO-LOAM和SegMatch的改进算法,阐述了该算法的原理和实现步骤;开展了煤矿井下模拟场景试验,对比分析改进前后SLAM算法的建图效果以及精度,试验结果表明改进算法构建的地图回环效果更好,估计轨迹更平滑、精确;结合导航需求研究了二维占据栅格地图的构建方法,试验验证了该方法所构建的栅格地图精度,结果表明有效滤除动态障碍物等离群噪点后的栅格地图具有0.01 m的建图精度,且所需存储空间较点云地图降低了3个数量级。研究成果有助于煤矿井下非结构环境下SLAM和煤矿机器人实时定位和自主移动。
