Research on unmanned driving system of underground trackless rubber-tyred vehicle in coal mine
-
摘要: 煤矿井下无轨胶轮车无人驾驶可大幅减少井下辅助运输作业人员数量，降低人员劳动强度，是辅助运输智能化的主要发展方向之一。相较于地面汽车无人驾驶，煤矿井下无轨胶轮车无人驾驶存在一系列新的挑战：井下巷道“长廊效应”、“多径效应”干扰；狭窄场景内人车混行等复杂路况对车辆精准控制的高要求；井下卫星拒止环境带来的定位问题；井下光照多变且巷道壁阻挡影响机器视觉的应用；设备需满足MA认证；安全措施需多重冗余设计等。针对上述挑战，提出了以车联网为核心的煤矿井下无轨胶轮车无人驾驶系统架构，分析了系统实现的关键技术：利用基于激光同步定位与建图（SLAM）和超宽带（UWB）/惯性导航系统（INS）的组合定位方式，实现车辆高速移动状态下的精确定位；依托车身多传感器（毫米波雷达、激光雷达、超声波雷达、摄像头）、矿用智能路侧单元等识别车身周边路况信息，并通过车联网共享相关信息；利用多源数据采集技术获得环境感知数据、车辆运行数据、路侧监控数据、移动目标数据，海量数据经5G等无线通信网络交互至基于边缘计算的分布式算力单元融合分析后，结合全局和局部路径规划算法合理规划车辆行驶路径，实现仓库管理系统化的车辆智能调度；考虑到井下机电设备安全准入要求，感知、线控、决策控制装备需实现矿用化设计且应尽量采用矿用本安型产品，以满足成本低、体积小、效率高的设计需求；井下无人驾驶车辆需实现感知、决策与控制环节的冗余设计，以实现非正常状况下车辆的安全可靠控制。现场测试结果表明：车辆定位精度可达0.3 m，通信带宽≥50 Mbit/s，数据通信时延≤50 ms，定位精度和数据交互满足井下无人驾驶基本需求；针对T形支巷及U型弯道等典型环境可实现避障及连续路径规划；基于多传感器融合策略，可实现多种目标感知能力提升；车辆动态跟随误差＜0.54 m/s，垂直于巷道壁方向平均控制误差＜0.2 m，满足无人驾驶车辆的控制要求。Abstract: The unmanned driving of underground trackless rubber-tyred vehicles in coal mine can significantly reduce the number of underground auxiliary transportation operating personnel, and reduce labor intensity. It is one of the leading development directions of intelligent auxiliary transportation. Compared with the unmanned driving of the ground vehicles, there are a series of new challenges for unmanned driving of underground trackless rubber-tyred vehicles. There is the interference of 'corridor effect' and 'multipath effect' in the underground roadway. There are high requirements for precise vehicle control under complex road conditions such as mixed traffic in narrow scenes. The underground satellite refusal environment causes positioning problems. Machine vision application is affected by the changeable illumination underground and the blocking of the roadway wall. The equipment shall meet MA certification. Multiple redundancy design is required for safety measures. In order to solve the above challenges, the architecture of the unmanned driving system for underground trackless rubber-tyred vehicle in coal mine based on the vehicle-to-everything is proposed. And the critical technologies of system implementation are analyzed. The integrated positioning method based on simultaneous localization and mapping (SLAM) and ultra wide band (UWB)/inertial navigation system (INS) is used to realize the precise positioning of the vehicle in the state of high-speed movement. By relying on the multi-sensor (millimeter-wave radar, laser radar, ultrasonic radar, camera) of the vehicle body and mining intelligent roadside unit, the road condition information around the vehicle body is identified. Through the vehicle-to-everything, the relevant information is shared. The multi-source data acquisition technology is used to obtain environmental perception data, vehicle operation data, roadside monitoring data, and mobile target data. The massive data is exchanged through 5G and other wireless communication networks to the distributed computing unit based on edge computing for fusion analysis. The vehicle driving path is reasonably planned in combination with global and local path planning algorithms to realize the systematic vehicle intelligent scheduling of warehouse management. Considering the safety access requirements of underground electromechanical equipment, the perception, wire control and decision-making control equipment shall be designed for mining. The mining intrinsically safe products shall be used as far as possible to meet the design requirements of low cost, small volume and high efficiency. Underground unmanned driving vehicles need to realize the redundant design of perception, decision-making and control links to realize the safe and reliable control of vehicles under abnormal conditions. The field test results show that the vehicle positioning precision can reach 0.3 m. The communication bandwidth is more than or equal to 50 Mbit/s. The data communication delay is less than or equal to 50 ms. Therefore the positioning precision and data exchange can meet the basic requirements of underground unmanned driving vehicles. The obstacle avoidance and continuous path planning can be realized for typical environments such as T-shaped roadway and U-shaped curve. Based on the multi-sensor fusion strategy, the perception capability of multiple targets can be improved. The vehicle dynamic following error is less than 0.54 m/s, and the average control error perpendicular to the roadway wall is less than 0.2 m. These results meet the control requirements of unmanned driving vehicles.
-
图 1 煤矿井下无轨胶轮车无人驾驶系统架构
Figure 1. Unmanned driving system architecture of trackless rubber-tyred vehicle in underground coal mine
图 2 基于多传感器融合的井下车辆路况识别
Figure 2. Road condition recognition of underground vehicle based on multi-sensor fusion
图 5 井下车辆智能调度工作机制
Figure 5. Working mechanism of intelligent dispatching of underground vehicle
图 6 矿用本安型高性能边缘计算装备电气原理
Figure 6. Electrical principle of mine intrinsically safe high-performance edge computing equipment
图 8 井下无人驾驶无轨胶轮车的权限管控及运转流程
Figure 8. Authorization control and operation process of underground unmanned driving trackless rubber-tyred vehicle
图 9 决策控制主机本安化设计方案
Figure 9. Intrinsically safety design scheme of decision control host
图 13 基于不同传感器的目标障碍物检测结果
Figure 13. Detection results of target obstacle by different sensors
图 14 无人驾驶车辆横纵向控制测试结果
Figure 14. Test results of lateral and longitudinal control of unmanned driving vehicle
表 1 室内定位技术对比
Table 1. Comparison of indoor positioning technologies
技术 精度/m 能耗 传输距离/m 抗干扰能力 优点 缺点 蓝牙 3.0 低 100 较弱 功耗低，穿透力较强 抗干扰能力较弱，定位精度较差 UWB 0.3 较低 250 强 抗干扰能力强，定位精度较高 需额外设备，部署较难 RFID 1.0～5.0 低 5 弱 功耗低，数据传输速率高 无法连续定位，定位精度差 ZigBee 3.0 较低 75 弱 成本及功耗较低 定位精度较差，抗干扰能力较弱 WiFi 5.0 较低 50 较弱 成本低，方便部署 定位精度差 激光 0.1 高 300 强 定位精度高，抗干扰能力强 功耗大，只可在可视范围内测距 超声波 1.0 较低 10 强 功耗较低，抗干扰能力强 有效定位距离短，部署难 
下载: 导出CSV
表 2 常用路径规划算法优缺点及适用环境
Table 2. Advantages, disadvantages and applicable environments of common path planning algorithms
算法 优点 缺点 适用环境 Dijkstra算法 使用贪心策略选择最优节点，能获得最优路径 运算过程中会占用大量计算资源，规划效率较低 环境信息已知的全局路径规划 A*算法 通过启发式采样方式搜索节点，算法搜索效率高 对地图要求较高，无法保证得到最优解 环境信息已知的全局路径规划 快速搜索随机树算法 通过随机采样搜索节点，适用于非完整约束场合 生成路径非最优，且规划效率低 环境信息已知的全局路径规划 人工势场法 能够实时避障，生成的路径平滑、安全 障碍物较多时容易陷入局部最优，造成目标不可达 既可用于全局路径规划，也可用于
局部路径规划动态窗口法 能够考虑无人驾驶车辆的速度与运动学约束，生成的路径平滑 存在局部最优解，且在复杂环境下计算复杂度高 环境信息部分已知的局部路径规划 蚁群算法 容易与其他启发式算法结合，能寻找到全局最优解 搜索具有盲目性，容易陷入局部最优 环境信息部分已知的局部路径规划
下载: 导出CSV
表 3 C−V2X通信时延及带宽
Table 3. Communication delay and bandwidth of C−V2X
交互模式 通信时延/ms 通信带宽/（Mbit${\boldsymbol{\cdot}} $s−1） V2N 上行 30～50 150～200 下行 20～30 50～100 V2I 上行 10～20 100～150 下行 5～15 50～80
下载: 导出CSV
表 4 井下目标障碍物感知统计结果
Table 4. Statistical results of underground target obstacles perception
目标类型 目标个数 准确率/% 毫米波
雷达激光
雷达摄像头 多传感器融合 大目标 390 96.3 95.9 93.7 98.9 中目标 350 94.2 91.2 93.2 98.6 小目标 300 86.5 83.5 89.9 96.5 极小目标 100 81.9 78.4 88.5 94.8
下载: 导出CSV
-
[1] 王国法. 煤矿智能化最新技术进展与问题探讨[J]. 煤炭科学技术,2022,50(1):1-27.WANG Guofa. New technological progress of coal mine intelligence and its problems[J]. Coal Science and Technology,2022,50(1):1-27. [2] 倪兴华. 安全高效矿井辅助运输关键技术研究与应用[J]. 煤炭学报,2010,35(11):1909-1915. doi: 10.13225/j.cnki.jccs.2010.11.027NI Xinghua. Research and application of key technology for safety and high efficient mine auxiliary transportation[J]. Journal of China Coal Society,2010,35(11):1909-1915. doi: 10.13225/j.cnki.jccs.2010.11.027 [3] 游小荣,裴浩,霍振龙. 一种基于UWB的三边定位改进算法[J]. 工矿自动化,2019,45(11):19-23. doi: 10.13272/j.issn.1671-251x.2019050081YOU Xiaorong,PEI Hao,HUO Zhenlong. An improved trilateral positioning algorithm based on UWB[J]. Industry and Mine Automation,2019,45(11):19-23. doi: 10.13272/j.issn.1671-251x.2019050081 [4] GITHINJI L. Effect of biochar application rate on soil physical and hydraulic properties of a sandy loam[J]. Archives of Agronomy and Soil Science,2014,60(4):457-470. doi: 10.1080/03650340.2013.821698 [5] SHAN Tixiao, ENGLOT B. LeGO-LOAM: lightweight and ground-optimized lidar odometry and mapping on variable terrain[C]//IEEE/RSJ International Conference on Intelligent Robots and Systems, Madrid, 2019: 4758-4765. [6] ASSEM I,DUPONT G. Friezes and a construction of the Euclidean cluster variables[J]. Journal of Pure and Applied Algebra,2011,215(10):2322-2340. doi: 10.1016/j.jpaa.2010.12.013 [7] LIU Yongfan,DU Sen,KONG Youyong. Supervoxel clustering with a novel 3D descriptor for brain tissue segmentation[J]. International Journal of Machine Learning and Computing,2020,10(3):501-506. doi: 10.18178/ijmlc.2020.10.3.964 [8] CHEN Yizhou,SHEN Shifei,CHEN Tao,et al. Path optimization study for vehicles evacuation based on Dijkstra algorithm[J]. Procedia Engineering,2014,71:159-165. doi: 10.1016/j.proeng.2014.04.023 [9] 鲍久圣,张牧野,葛世荣,等. 基于改进A*和人工势场算法的无轨胶轮车井下无人驾驶路径规划[J]. 煤炭学报,2022,47(3):1347-1360. doi: 10.13225/j.cnki.jccs.xr21.1716BAO Jiusheng,ZHANG Muye,GE Shirong,et al. Underground driverless path planning of trackless rubber tyred vehicle based on improved A* and artificial potential field algorithm[J]. Journal of China Coal Society,2022,47(3):1347-1360. doi: 10.13225/j.cnki.jccs.xr21.1716 [10] ZHANG Zhenghao,QIAO Bing,ZHAO Wentong,et al. A predictive path planning algorithm for mobile robot in dynamic environments based on rapidly exploring random tree[J]. Arabian Journal for Science and Engineering,2021,46(9):8223-8232. doi: 10.1007/s13369-021-05443-8 [11] 田子建,高学浩,张梦霞. 基于改进人工势场的矿井导航装置路径规划[J]. 煤炭学报,2016,41(增刊2):589-597. doi: 10.13225/j.cnki.jccs.2016.1165TIAN Zijian,GAO Xuehao,ZHANG Mengxia. Path planning based on the improved artificial potential field of coal mine dynamic target navigation[J]. Journal of China Coal Society,2016,41(S2):589-597. doi: 10.13225/j.cnki.jccs.2016.1165 [12] 袁晓明,郝明锐. 煤矿无轨辅助运输无人驾驶关键技术与发展趋势研究[J]. 智能矿山,2020,1(1):89-97.YUAN Xiaoming,HAO Mingrui. Key technology and development trend of mine auxiliary transport autonomous vehicle[J]. Journal of Intelligent Mine,2020,1(1):89-97. [13] 谭玉新,杨维,徐子睿. 面向煤矿井下局部复杂空间的机器人三维路径规划方法[J]. 煤炭学报,2017,42(6):1634-1642. doi: 10.13225/j.cnki.jccs.2016.1047TAN Yuxin,YANG Wei,XU Zirui. Three-dimensional path planning method for robot in underground local complex space[J]. Journal of China Coal Society,2017,42(6):1634-1642. doi: 10.13225/j.cnki.jccs.2016.1047 [14] LIU Jianhua,YANG Jianguo,LIU Huaping,et al. An improved ant colony algorithm for robot path planning[J]. Soft Computing,2017,21(19):5829-5839. doi: 10.1007/s00500-016-2161-7 [15] 张朝阳. 矿用无轨胶轮车无人驾驶系统研究[D]. 西安: 西安科技大学, 2016.ZHANG Chaoyang. Research on unmanned system for mine trackless rubber wheel vehicle[D]. Xi'an: Xi'an University of Science and Technology, 2016. [16] 周晶晶, 苏致远, 马育林, 等. 基于多传感器的智能车交通状态感知关键技术研究[C]//第11届中国智能交通年会大会, 重庆, 2016: 688-692.ZHOU Jingjing, SU Zhiyuan, MA Yulin, et al. Research on key technology of intelligent vehicle traffic state perception based on multi-sensor[C]//The 11th China Intelligent Transportation Annual Conference, Chongqing, 2016: 688-692. [17] 任大凯,廖振松. 5G车路协同自动驾驶应用研究[J]. 电信工程技术与标准化,2020,33(9):68-74. doi: 10.3969/j.issn.1008-5599.2020.09.014REN Dakai,LIAO Zhensong. Research on application of 5G-V2X autonomous driving[J]. Telecom Engineering Technics and Standardization,2020,33(9):68-74. doi: 10.3969/j.issn.1008-5599.2020.09.014 [18] 王斌. 煤矿无轨辅助运输设备的应用与发展趋势[J]. 煤矿机械,2013,34(8):1-3. doi: 10.13436/j.mkjx.2013.08.117WANG Bin. Application and development of coal mine trackless auxiliary transportation equipment[J]. Coal Mine Machinery,2013,34(8):1-3. doi: 10.13436/j.mkjx.2013.08.117 [19] 刘宏杰,张慧,张喜麟,等. 煤矿无轨胶轮车智能调度管理技术研究与应用[J]. 煤炭科学技术,2019,47(3):81-86. doi: 10.13199/j.cnki.cst.2019.03.011LIU Hongjie,ZHANG Hui,ZHANG Xilin,et al. Research and application of intelligent dispatching and management technology for coal mine trackless rubber-tyred vehicle[J]. Coal Science and Technology,2019,47(3):81-86. doi: 10.13199/j.cnki.cst.2019.03.011 [20] 李建明. 梅花井煤矿辅助运输系统选择及应用研究[J]. 煤炭科技,2014,35(3):1-2. doi: 10.3969/j.issn.1008-3731.2014.03.002LI Jianming. Research on selection and application of auxiliary transportation system in Meihuajing Coal Mine[J]. Coal Science & Technology Magazine,2014,35(3):1-2. doi: 10.3969/j.issn.1008-3731.2014.03.002 [21] 吴建波,朱文霞,剧亮,等. 边缘计算在智慧交通系统中的应用[J]. 计算机与现代化,2021(12):103-109,122. doi: 10.3969/j.issn.1006-2475.2021.12.017WU Jianbo,ZHU Wenxia,JU Liang,et al. Application of edge computing in intelligent transportation systems[J]. Computer and Modernization,2021(12):103-109,122. doi: 10.3969/j.issn.1006-2475.2021.12.017 [22] 陈霄,刘巍,陈静,等. 边缘计算环境下的计算卸载策略研究[J]. 火力与指挥控制,2022,47(1):7-14,19. doi: 10.3969/j.issn.1002-0640.2022.01.002CHEN Xiao,LIU Wei,CHEN Jing,et al. Research on computing offload strategy in edge computing environment[J]. Fire Control & Command Control,2022,47(1):7-14,19. doi: 10.3969/j.issn.1002-0640.2022.01.002 [23] 杨晓丹. 煤矿井下防爆电气设备中的应用技术[J]. 电子技术与软件工程,2019(24):223-224.YANG Xiaodan. Application technology of explosion-proof electrical equipment in coal mine[J]. Electronic Technology & Software Engineering,2019(24):223-224. [24] 陈山枝,时岩,胡金玲. 蜂窝车联网(C−V2X)综述[J]. 中国科学基金,2020,34(2):179-185. doi: 10.16262/j.cnki.1000-8217.2020.02.009CHEN Shanzhi,SHI Yan,HU Jinling. Cellular vehicle to everything(C-V2X):a review[J]. Bulletin of National Natural Science Foundation of China,2020,34(2):179-185. doi: 10.16262/j.cnki.1000-8217.2020.02.009 [25] 陈山枝,葛雨明,时岩. 蜂窝车联网(C−V2X)技术发展、应用及展望[J]. 电信科学,2022,38(1):1-12.CHEN Shanzhi,GE Yuming,SHI Yan. Technology development,application and prospect of cellular vehicle-to-everything(C-V2X)[J]. Telecommunications Science,2022,38(1):1-12. [26] 阎俊豪,贾宗璞,李东印. 智能矿山车联网体系架构与关键技术[J]. 煤炭科学技术,2020,48(7):249-254. doi: 10.13199/j.cnki.cst.2020.07.026YAN Junhao,JIA Zongpu,LI Dongyin. Architecture and key technologies of intelligent of vehicles in intelligent mine[J]. Coal Science and Technology,2020,48(7):249-254. doi: 10.13199/j.cnki.cst.2020.07.026 [27] 韩江洪,卫星,陆阳,等. 煤矿井下机车无人驾驶系统关键技术[J]. 煤炭学报,2020,45(6):2104-2115. doi: 10.13225/j.cnki.jccs.ZN20.0343HAN Jianghong,WEI Xing,LU Yang,et al. Driverless technology of underground locomotive in coal mine[J]. Journal of China Coal Society,2020,45(6):2104-2115. doi: 10.13225/j.cnki.jccs.ZN20.0343 [28] 闫凌,黄佳德. 矿用卡车无人驾驶系统研究[J]. 工矿自动化,2021,47(4):19-29. doi: 10.13272/j.issn.1671-251x.17729YAN Ling,HUANG Jiade. Research on unmanned driving system of mine-used truck[J]. Industry and Mine Automation,2021,47(4):19-29. doi: 10.13272/j.issn.1671-251x.17729 [29] 于月森,谢冬莹,李世光,等. 本质安全电路技术综述[J]. 煤炭科学技术,2011,39(6):61-65. doi: 10.13199/j.cst.2011.06.67.yuys.025YU Yuesen,XIE Dongying,LI Shiguang,et al. Summary of intrinsic safety electric circuit technology[J]. Coal Science and Technology,2011,39(6):61-65. doi: 10.13199/j.cst.2011.06.67.yuys.025 [30] 林引. 矿用高可靠性本安型传感器电源电路设计与实现[J]. 煤炭科学技术,2013,41(6):88-91.LIN Yin. Design and realization on power of high reliable intrinsic safe sensor[J]. Coal Science and Technology,2013,41(6):88-91. [31] 王璇. 矿用本安型网口电路设计[J]. 煤矿安全,2016,47(6):113-114,118. doi: 10.13347/j.cnki.mkaq.2016.06.031WANG Xuan. Design of mine-used intrinsic safe network interface circuit[J]. Safety in Coal Mines,2016,47(6):113-114,118. doi: 10.13347/j.cnki.mkaq.2016.06.031 -
煤矿井下无轨胶轮车无人驾驶系统研究+增强视频.mp4
-
点击查看大图
图(14) / 表(4)
计量
- 文章访问数: 1427
- HTML全文浏览量: 110
- PDF下载量: 153
- 被引次数: 0
