随着综采工作面自动化程度迅速提高，掘进效率不足导致的采掘失衡成为目前煤炭开采面临主要问题。悬臂式掘进机具有投资少、施工准备时间短，可用于任何断面形状掘进等优势，是硬度较低的全岩巷道和半煤岩巷道掘进的主要方式。悬臂式掘进机自动化、智能化发展对提高煤炭资源的高效开采具有重要意义。

论文针对悬臂式掘进机断面自动成形截割和定向掘进控制进行研究，通过对比目前国内外悬臂式掘进机位姿测量方法，提出一种基于单目视觉测量的悬臂式掘进机位姿测量与自动控制方法，结合多传感器实现掘进机截割头和机身的位姿测量，构建基于视觉伺服的悬臂式掘进机自动控制系统，实现巷道掘进的定位、定向和定形控制。

针对煤矿井下设备精确定位难题，提出一种基于机器视觉的悬臂式掘进机机身及截割头位姿测量方法。采用单目视觉测量技术，以多点红外LED标靶和多束激光指向仪为目标特征，利用掘进机机身安装的防爆相机采集图像并进行一系列算法处理获得掘进机器人机身及截割头的位姿坐标。论文对悬臂式掘进机运动学模型进行分析，构建系统全局坐标系，建立基于共面特征点的截割头位姿测量模型和基于门形三线的机身位姿测量模型，进一步解算得到巷道坐标系中悬臂式掘进机截割头和机身位姿。

针对巷道非全断面成形截割控制智能化程度低的问题，将视觉伺服技术应用于悬臂式掘进机截割头控制，提出基于视觉伺服的煤矿巷道非全断面成形截割控制方法。以多点红外LED标靶解决高粉尘、低照度井下环境的图像获取问题，实时可靠地获取截割头位姿测量数据，按照非全断面施工工艺规划的截割轨迹路径，以截割头实时位姿为反馈量，构建基于PID的视觉伺服截割头轨迹跟踪控制模型，实现非全断面巷道成形截割。

针对煤矿巷道定向掘进施工质量控制难题，提出一种基于精确位姿测量和轨迹跟踪控制的悬臂式掘进机定向掘进方法。研究了井下高粉尘、多杂光环境下激光点、线特征提取方法，结合多传感器数据实现悬臂式掘进机的位姿测量。通过对悬臂式掘进机行走机构运动学分析，构建巷道施工中的悬臂式掘进机定向掘进模型，采用滑模变结构方法设计轨迹跟踪控制器，实现巷道定向掘进，通过仿真验证算法的有效性。

最后，搭建悬臂式掘进机监控系统，实现数据采集、图像处理、位姿解算以及掘进机在线监控等功能，在实验室以及煤矿井下环境中完成各模块的功能验证。实验结果表明论文提出的悬臂式掘进机位姿测量和智能控制精度均达到设计要求，煤矿井下掘进验证了系统的可行性，证明基于视觉测量构建的掘进装备位姿检测系统有助于解决巷道掘进的自动化难题，结合视觉伺服和轨迹跟踪控制等方法可以保证巷道掘进精度，满足巷道施工的质量要求。

With the rapid improvement of the automation of the fully mechanized coal face equipment, the unbalance between mining and heading caused by the low driving efficiency has become the main problem of coal mining. Boom-type roadheader has the advantages of less investment, short construction preparation time, and been used in any shape of cross-section, etc. It is the main way of driving in full rock roadway and half coal rock roadway with low hardness. The automation and intelligent development of boom-type roadheaders are of great significance for improving the efficient mining of coal resources.

This paper studies the automatic forming and cutting of the boom-type roadheader section and the directional tunneling control. By comparing the current domestic and foreign methods of the boom-type roadheader position and attitude measurement, a method of the boom-type roadheader position and attitude measurement and automatic control based on the visual measurement is proposed. Combined with the multi-sensor, the position and attitude measurement of the cutting head and the body of the roadheader is realized, and the suspension based on the visual servo is constructed The automatic control system of the boom-type roadheader realizes the positioning, orientation and shape control of the roadheader.

Aiming at the problem of precise positioning of underground equipment in coal mine, this paper presents a method of measuring the position and orientation of the body and cutting head of the boom-type roadheader based on machine vision. Taking multi-point infrared LED target and multi beam laser pointer as the target characteristics, using the explosion-proof camera installed on the body of the roadheader to collect images and carry out a series of algorithm processing to obtain the position and orientation coordinates of the body and cutting head of the roadheader robot. This paper analyzes the kinematics model of roadheader, constructs the global coordinate system of the system, establishes the cutting head pose measurement model based on coplanar feature points and the fuselage pose measurement model based on the three-lines of the gate shape, and further calculates the cutting head and fuselage pose of boom-type roadheader in the tunnel coordinate system.

In view of the low intelligent degree of non full section cutting control, the visual servo technology is applied to the cutting head control of boom-type roadheader, and the control method of non full section cutting based on visual servo is proposed. The multi-point infrared LED target is used to solve the problem of image acquisition in the underground environment with high dust and low illumination. The real-time and reliable measurement data of cutting head's position and attitude are obtained. According to the cutting path of non full section construction process planning, the real-time position and attitude of cutting head is used as the feedback quantity to construct the visual servo tracking control model of cutting head based on PID, so as to realize the forming cutting of non full section roadway.

In order to solve the quality control problem of directional driving construction in coal mine, a method of directional driving of the boom-type roadheader based on precise position and attitude measurement and track tracking control is proposed. This paper studies the feature extraction method of laser point and line under the environment of high dust and multi light, and realizes the position and attitude measurement of the boom-type roadheader with multi-sensor data. Based on the kinematic analysis of the walking mechanism of the boom-type roadheader, the directional driving model of the boom-type roadheader in the tunnel construction is constructed, and the track tracking controller is designed by the sliding mode variable structure method to realize the directional driving of the tunnel, and the effectiveness of the algorithm is verified through simulation.

Finally, Set up the monitoring system of boom-type roadheader to realize the functions of data acquisition, image processing, pose calculation and online monitoring of the roadheader. The function verification of each module is completed in the laboratory and underground environment of the coal mine. The experimental results show that the position and attitude measurement and intelligent control accuracy of the boom-type roadheader proposed in this paper meet the design requirements, which verifies the feasibility of the system. It is proved that the position and attitude detection system of the driving equipment based on the visual measurement is helpful to solve the automatic problem of the tunnel driving. The combination of visual servo and track tracking control can ensure the accuracy of the tunnel driving, Meet the quality requirements of roadway construction.