在当前煤炭行业的高质量发展背景下，智能化掘进装备的研发和应用日益成为提高生产效率和安全性的核心目标。其中，位姿检测技术作为智能化掘进装备的重要组成部分，其精准度直接影响到装备的定位和导航性能，因此受到了广泛的关注。然而，在实际应用中，由于井下环境的复杂性和掘进装备的连续移动特性，存在因环境干扰导致相机易受遮挡以及视觉测量数据不稳定的问题。且现有基于视觉的位姿测量方案中，在合作标靶移站后的标定过程仍存在一定的局限性，难以满足快速掘进的需求。针对上述问题，提出了一种煤矿井下双单目视觉定位系统与自动标定技术，旨在解决井下巷道掘进装备视觉定位系统在快速掘进过程中的连续移站测量问题。主要研究内容如下：

针对使用单目视觉方法进行井下掘进装备定位时，存在因环境干扰导致相机易受遮挡以及视觉测量数据不稳定的问题，提出了一种基于双单目视觉的掘进装备定位方法。通过分析双单目定位原理，结合扩展卡尔曼滤波和粒子滤波算法，构建多源传感器融合算法，实现了煤矿井下掘进装备的双单目视觉位姿融合，得到了掘进装备在巷道坐标系下的最优位姿。该方法有效提升了视觉定位系统的抗遮挡性与鲁棒性。

针对掘进装备视觉定位系统在快速掘进过程中的连续移站测量问题，提出了一种基于双单目视觉的移站后自动标定方法。针对合作标靶点线特征的预处理流程，结合井下巷道的特殊环境，采用基于HSV（Hue, Saturation, Value）颜色空间分割技术和特征提取技术。通过自适应轮廓检测和椭圆拟合方法精确获取光斑中心坐标，并结合梯度外接矩形和Hough变换确定激光束的中心线。利用双单目立体模型和Levenberg-Marquardt（LM）算法，解算出合作标靶的空间点线特征信息，为双单目视觉测量提供数据支撑。

针对掘进装备在快速掘进过程中，视觉定位系统建站移站过程中存在的问题，研究掘进装备视觉定位系统建站移站性能。通过分析双单目视觉定位过程中的影响因素，建立双单目定位过程中的误差理论模型；同时，对双相机空间结构进行了详细分析，建立了双单目移站后标定过程中的误差模型。最后，将双单目定位和双单目移站后标定过程的误差模型进行结合，对建站移站过程的总误差进行了全面分析。为后续相关研究提供了重要的理论基础和参考依据。

在上述理论和方法研究的基础上，搭建实验平台，验证上述方法的可行性。结果表明，在保证较好的位姿测量精度下，移站后标定方法能够有效的增强掘进装备视觉位姿测量系统的测量距离，并且在建站移站过程中，双单目视觉定位系统与自动标定技术可以提供稳定有效的机身位姿数据。本研究旨在提供一种掘进装备视觉测量新思路以及自动标定方法，解决掘进装备在井下复杂环境中的定位和标定难题，为快速掘进技术以及自主截割技术的研究提供重要支持。

In the context of high-quality development in the current coal industry, the research and application of intelligent tunneling equipment have increasingly become core goals for enhancing production efficiency and safety. Pose detection technology, as an essential component of intelligent tunneling equipment, directly impacts the equipment's positioning and navigation performance, thus receiving widespread attention. However, in practical applications, the complexity of the underground environment and the continuous movement of tunneling equipment lead to challenges such as cameras being easily occluded by environmental interference and instability in visual measurement data. Moreover, existing vision-based pose measurement schemes face limitations in the calibration process after cooperative target relocation, which hinders the needs for rapid tunneling. To address these issues, a dual monocular vision positioning system and automatic calibration technology for underground coal mines is proposed. The aim is to solve the problem of continuous station movement measurements in the visual positioning system of tunneling equipment during rapid tunneling. The main research contents are as follows:

Aiming at the issues of camera occlusion and unstable visual measurement data when using monocular vision methods for underground excavation equipment positioning, a method based on dual-monocular vision is proposed. By analyzing the principles of dual-monocular positioning and combining extended Kalman filtering and particle filtering algorithms, a multi-source sensor fusion algorithm is constructed to achieve the optimal pose fusion of excavation equipment under the coordinate system of the roadway. This method effectively enhances the occlusion resistance and robustness of the visual positioning system.

Addressing the problem of continuous station measurement of excavation equipment visual positioning systems during rapid excavation, a method based on binocular vision for autonomous station calibration is proposed. Pre-processing procedures for cooperative target point-line features, combined with the special environment of underground roadways, employ techniques such as HSV (Hue, Saturation, Value) color space segmentation and feature extraction. Accurate acquisition of the centroid coordinates of light spots is achieved through adaptive contour detection and ellipse fitting methods, combined with gradient bounding rectangles and Hough transforms to determine the centerline of laser beams. Utilizing binocular stereo models and the Levenberg-Marquardt (LM) algorithm, spatial point-line feature information of cooperative targets is calculated, providing data support for dual-monocular visual measurements.

Studying the performance of excavation equipment visual positioning systems during station building and moving processes. By analyzing influencing factors in the dual-monocular visual positioning process, error theoretical models for the dual-monocular positioning process are established. Simultaneously, detailed analysis of the dual-camera spatial structure is conducted, establishing error models for binocular station calibration processes. Finally, the error models of the dual-monocular positioning and binocular station calibration processes are combined to comprehensively analyze the total error of the station building and moving processes, providing important theoretical foundations and reference for subsequent related research.

Based on the above theoretical and methodological research, an experimental platform is constructed to validate the feasibility of the proposed methods. The results show that, while ensuring good pose measurement accuracy, the station calibration method effectively enhances the measurement distance of the excavation equipment visual pose measurement system. Moreover, during the station building and moving processes, the dual-monocular visual positioning system and automatic calibration technology can provide stable and effective body pose data. This study aims to provide a new approach to visual measurement of excavation equipment and automatic calibration methods, solving the positioning and calibration challenges of excavation equipment in complex underground environments, and providing essential support for research on rapid excavation technology and autonomous cutting technology.

[1]杨沐岩.煤炭供应总量再创新高开发布局持续优化[N].中国能源报,2024-04-01(003).

[2]张翔.量变质升：2023煤炭工业发展主线[N].中国煤炭报,2024-03-30(001).

[3]王琼杰.夯实煤炭主体能源根基[N].中国矿业报,2024-03-05(002).

[4]Shi-Rong G E. Key technology of intelligent coal mining equipment[J]. Coal Science and Technology, 2014 (9).

[5]张旭辉,刘博兴,张超,等.掘进装备全站仪与捷联惯导组合定位方法[J].工矿自动化, 2020, 46(09): 1-7.

[6]Wu G, Fang X, Zhang L, et al. Positioning accuracy of the shearer based on a strapdown inertial navigation system in underground coal mining[J]. Applied Sciences, 2020, 10(6): 2176.

[7]Han R, Dunker T, Trostmann E, et al. Research on High-Precision Tracking and Localization of Industrial Robots Based on iGPS[J]. Measurement Science and Technology, 2024.

[8]Kang X, Li J, Fan X, et al. Real-time rgb-d simultaneous localization and mapping guided by terrestrial lidar point cloud for indoor 3-d reconstruction and camera pose estimation[J]. Applied Sciences, 2019, 9(16): 3264.

[9]张蒙恩.浅谈全断面掘进机(TBM)激光导向系统[J].水电站机电技术,2016,39(S1):89-92.

[10]Pham V T, Nguyen D C, Tran Q H, et al. Thermal stability of magnetic compass sensor for high accuracy positioning applications[J]. Sensors & Transducers, 2015, 195(12): 1.

[11]Fu S, Li Y, Zong K, et al. Ultra-wideband pose detection method based on TDOA positioning model for boom-type roadheader[J]. AEU-International Journal of Electronics and Communications, 2019, 99: 70-80.

[12]杨文娟,张旭辉,张超,等.基于三激光束标靶的煤矿井下长距离视觉定位方法[J].煤炭学报, 2022, 47(02): 986-1001.

[13]胡兴涛,朱涛,苏继敏,等.煤矿巷道智能化掘进感知关键技术[J].煤炭学报, 2021, 46(07): 2123-2135．

[14]杨秀宇.基于光电传感器的掘进机位姿测量方法[J].机械设计, 2022,39(07):117-121.

[15]程健.煤矿巷道机器人管线视觉辅助定位与导航方法研究[J].煤炭科学技术, 2020, 48(07): 226-232.

[16]张旭辉,沈奇峰,杨文娟,等.基于三激光点标靶的掘进机机身视觉定位技术研究[J].电子测量与仪器学报,2022,36(06):178-186.

[17]杜雨馨,刘停,童敏明,等.基于机器视觉的悬臂式掘进机机身位姿检测系统[J].煤炭学报, 2016, 41(11): 2897-2906.

[18]Yu R, Fang X, Hu C, et al. Research on positioning method of coal mine mining equipment based on monocular vision[J]. Energies, 2022, 15(21): 8068.

[19]田原.悬臂式掘进机视觉定位方法研究[J].矿山机械, 2019, 47(03): 8-12.

[20]薛旭升, 张旭辉, 毛清华, 等. 基于双目视觉的掘进机器人定位定向方法研究[J]. 西安科技大学学报, 2020, 40(05): 781-789.

[21]王学,周红旭,张雷,等.基于近红外双目立体视觉的悬臂式掘进机定位研究[J].工矿自动化, 2022, 48(07): 43-57.

[22]Wan J, Zhang X, Zhang C, et al. Vision and Inertial Navigation Combined-Based Pose Measurement Method of Cantilever Roadheader[J]. Sustainability, 2023, 15(5): 4018.

[23]张超.悬臂式掘进机双目视觉与捷联惯导组合定位技术研究[D].西安:西安科技大学, 2020.

[24]黄东,杨凌辉,罗文,等.基于视觉/惯导的掘进机实时位姿测量方法研究[J].激光技术, 2017, 41(01): 19-23.

[25]王国法,杜毅博,陈晓晶,等.从煤矿机械化到自动化和智能化的发展与创新实践——纪念《工矿自动化》创刊50周年[J]. 工矿自动化,2023,49(06):1-18.

[26]杜雨馨,张贺,王树臣,张建化.综掘系统视觉处理技术研究现状及发展趋势[J].工矿自动化,2023,49(11):22-38+75.

[27]田臣,覃杰,王强,唐会成.煤矿智能掘进装备发展现状与关键技术探讨[J].中国煤炭,2023,49(7):102-108.

[28]田伟琴,田原,贾曲,张凯.悬臂式掘进机导航技术研究现状及发展趋势[J].煤炭科学技术,2022,50(03):267-274.

[29]胡兴涛,朱涛,苏继敏,等.煤矿巷道智能化掘进感知关键技术[J].煤炭学报,2021,46(7): 023-2135.

[30]杨健健,张强,吴淼,等.巷道智能化掘进的自主感知及调控技术研究进展[J].煤炭学报,2020,45(6):2045-2055.

[31]张旭辉,张楷鑫,张超,等.悬臂式掘进机视觉位姿检测系统外参标定方法[J].机械科学与技术,2022,41(12):1928-1935.

[32]Zou X, Zou H, Lu J. Virtual manipulator-based binocular stereo vision positioning system and errors modelling[J]. Machine Vision and Applications, 2012, 23: 43-63.

[33]Ma W P, Li W X, Cao P X. Binocular vision object positioning method for robots based on coarse-fine stereo matching[J]. International Journal of Automation and Computing, 2020, 17: 562-571.

[34]Lu K, Wang X, Wang Z, et al. Binocular stereo vision based on OpenCV[J]. 2011.

[35]Lai X, Wang H, Xu Y. A real-time range finding system with binocular stereo vision[J]. International Journal of Advanced Robotic Systems, 2012, 9(1): 26.

[36]Zheng J, Ding F. A filtering‐based recursive extended least squares algorithm and its convergence for finite impulse response moving average systems[J]. International Journal of Robust and Nonlinear Control, 2024.

[37]Zhong S, Peng B, He J, et al. Kalman filtering based on dynamic perception of measurement noise[J]. Mechanical Systems and Signal Processing, 2024, 213: 111343.

[38]Choi A, Ahn H, Chung Y, et al. Sliding Mode Control for Sensorless Speed Tracking of PMSM with Whale Optimization Algorithm and Extended Kalman Filter[J]. Machines, 2023, 11(9): 851.

[39]Choppala P B. Auxiliary particle filtering with lookahead support for univariate state space models[J]. ANNALI DELL'UNIVERSITA'DI FERRARA, 2024: 1-18.

[40]陈健,孟义兴,袁慎芳,等.融合导波监测的搭接结构裂纹扩展寿命孪生预测[J/OL].机械工程学报:1-9[2024-04-25].

[41]He J. Research on Mobile Node Localization Optimization Algorithm in Wireless Sensor Networks Based on Monte Carlo Method[J]. Wireless Communications and Mobile Computing, 2022, 2022(1): 7131099.

[42]李玉刚,陈克,董振飞.基于多传感器融合的目标检测方法研究[J].内燃机与配件,2024,(07):31-33.

[43]Bar-Shalom Y. On the track-to-track correlation problem[J]. IEEE Transactions on Automatic control, 1981, 26(2): 571-572.

[44]曾钰桔.基于融合算法的AGV路径规划及避障研究[D].昆明理工大学,2023.

[45]徐万金.基于卡尔曼滤波和粒子滤波的多信息融合目标跟踪[D].兰州交通大学,2024.

[46]付学瀚,燕贺云,朱立东,等.基于卡尔曼和扩展卡尔曼滤波的耦合载波跟踪方法[J].移动通信,2024,48(01):118-124.

[47]彭道刚,段睿杰,王丹豪.两级融合的多传感器数据融合算法研究[J].仪表技术与传感器,2024(01):87-93.

[48]阳兆哲,李跃忠,吴光文.基于无迹卡尔曼滤波和小波分析的IMU传感器去噪技术研究[J].现代电子技术,2024,47(05):53-59.

[49]毛清华,周庆,安炎基,等.惯导与视觉信息融合的掘进机精确定位方法[J/OL].煤炭科学技术:1-17[2024-04-15].

[50]高梓.悬臂式掘进机组合定位导航方案设计与试验[J].山西焦煤科技,2023,47(08):4-7.

[51]刘送永,崔玉明,孟德远,等.巷道掘进机多传感融合定位系统及试验研究[J].振动.测试与诊断,2023,43(03):476-484+618.

[52]Wang H, Li Z, Wang H, et al. A Roadheader Positioning Method Based on Multi-Sensor Fusion[J]. Electronics, 2023, 12(22): 4556.

[53]Cui Y, Liu S, Li H, et al. Accurate integrated position measurement system for mobile applications in GPS-denied coal mine[J]. ISA transactions, 2023, 139: 621-634.

[54]吴雷,孙永荣,李旺灵等.自主空中加油单目视觉导航误差影响因素研究[J].电子测量与仪器学报,2018,32(11):132-138.

[55]郝仁杰,王中宇,李亚茹.一种单目视觉位姿测量系统的误差分析方法[J].应用光学,2019,40(01):79-85.

[56]陈思伟.视觉引导抓取机械手工作平面定位误差与修正[D].天津大学,2015.

[57]张全喜.单目视觉自主移动机器人的目标跟踪及误差分析[D].南京理工大学,2014.

[58]毕超,郝雪,鄂玛兰,等.基于误差模型的双目视觉系统测量精度分析[J].航空精密制造技术,2020,56(02):1-4+30.

[59]王龙华,曾国华,胡华锋,等.精确三维数据测量的双目视觉系统优化设计[J].电子测量技术,2022,45(07):99-109.

[60]张红伟,朱运东,林雪竹,等.双目视觉测量网络布局规划技术研究[J].长春理工大学学报(自然科学版),2021,44(02):27-36+42.

[61]杨丽丽,王华涛.基于双目视觉的高精度摄影系统测量精度模型分析[J].工程与试验,2018,58(01):7-10+85.

[62]孟环标.双目视觉测量系统误差分析与控制[D].山东大学,2013.

[63]邸男,田睿.空间机器人双目视觉测量系统精度分析[J].载人航天,2017,23(01):76-81.

[64]Liu G, Zheng Q, Niu S, et al. Research and application of the global positioning system (GPS) clustering algorithm based on multilevel functions[J]. Journal of Computational Methods in Sciences and Engineering, 2024, 24(1): 357-368.

[65]刘仲民,李战明.基于模糊神经网络的双目视觉定位系统研究[J].计算机工程与设计, 2010, 31(03): 599-601.

[66]Khang A ,Abdullayev V ,Misra A , et al.Machine Vision and Industrial Robotics in Manufacturing:Approaches, Technologies, and Applications[M].CRC Press:2024-03-30.

[67]张杜娟.基于肤色模型与颜色空间的目标识别算法[J].国外电子测量技术,2019,38(10):54-58.

[68]王宁,宋慧慧,张开华.基于距离加权重叠度估计与椭圆拟合优化的精确目标跟踪算法[J].计算机应用,2021,41(4):1100-1105.

[69]刘爱胤,熊根良,姚健康,等.基于霍夫圆检测的标志物多图像特征定位方法[J].计算机仿真,2021,38(3):341-345.

[70]王学敏,于洪波,张翔宇,等.基于Hough变换检测前跟踪的水下多目标被动检测方法[J]. 兵工学报,2023,44(07):2114-2121.

[71]马颂德,张正友.计算机视觉-计算理论与算法基础[M].科学出版社.1998.52-73.

[72]Ahmadi S P, Hansson A, Pakazad S K. Distributed localization using Levenberg-Marquardt algorithm[J]. EURASIP Journal on Advances in Signal Processing, 2021, 2021: 1-26.

[73]Gupta A, Kumar S. A multiplicative Gauss-Newton minimization algorithm: Theory and application to exponential functions[J]. Applied Mathematics-A Journal of Chinese Universities, 2021, 36: 370-389.