College of Mechanical Engineering, Xi’an University of Science and Technology
Shaanxi Key Laboratory of Mine Electromechanical Equipment Intelligent Monitoring
Xi’an Coal Mining Machinery Co., Ltd.
Shaanxi Mincetech Co., Ltd.

针对当前掘进设备智能化程度低，掘进过程中超挖、欠挖和异常碰撞等情况难以具体描述，传统自动截割和记忆截割技术难以适应复杂地质条件等问题，提出了一种数字孪生驱动的悬臂式掘进机虚拟示教记忆截割控制方法。通过分析数字孪生技术在当前煤炭智能化开采领域的研究情况，设计了数字孪生驱动的悬臂式掘进机记忆截割控制系统总体方案，对复杂工况下悬臂式掘进机记忆截割关键技术进行研究。首先，充分利用数字孪生和虚拟现实技术的特性，研究复杂工况环境下的虚拟示教策略，基于Unity3D平台建立对应实体等尺寸的工作面及设备虚拟孪生模型、截割部运动学模型和虚拟碰撞检测模型，在虚拟端通过智能交互界面控制虚拟模型运动，结合工人经验设计示教轨迹并优化，使其作为轨迹跟踪的目标期望轨迹，弥补传统井下人工示教由于恶劣工况环境导致的过度依赖工人经验等问题；其次，为提高断面成形质量，研究记忆自动截割阶段示教轨迹跟踪再现控制方法，利用拉格朗日法建立截割部动力学模型，采用迭代学习与滑模控制相结合的方法提高末端执行器对示教轨迹的跟踪控制精度；最后，搭建悬臂式掘进机记忆截割仿真控制平台，通过虚拟空间与物理空间以及各模块之间的实时数据传输与交互，在虚拟空间中完成记忆截割虚拟示教与轨迹跟踪控制过程的三维可视化仿真，进而生成记忆自动截割轨迹跟踪控制指令并发送至悬臂式掘进机物理实体的末端执行器，驱动其按照示教轨迹进行断面成形截割，同时物理传感器采集悬臂式掘进机机身及截割臂位姿数据，反向驱动虚拟模型同步动作，实现对机器人虚拟模型和物理实体的闭环控制。在此基础上对系统的虚实同步性、虚拟样机与物理样机运动一致性和轨迹跟踪再现控制精度进行验证，实验结果表明，系统数据传输延迟较低，能够保证虚实一致性和同步性，且轨迹跟踪控制精度满足实际使用要求，该方法为掘进设备记忆截割与智能化控制提供了新的思路。

Aiming at the problems of low intelligence of current tunneling equipment, difficulty in describing over-excavation, under-excavation and abnormal collision in tunneling process, and difficulty in adapting traditional automatic cutting and memory cutting technology to complex geological conditions, a digital twin-driven virtual teaching memory cutting control method for cantilever roadheader is proposed. By analyzing the research situation of digital twin technology in the field of intelligent coal mining, the overall scheme of memory cutting control system of cantilever roadheader driven by digital twin is designed, and the key technology of memory cutting of cantilever roadheader under complex working conditions is studied. Firstly, the characteristics of digital twin and virtual reality technology are fully utilized to study the virtual teaching strategy under complex working conditions. Based on the Unity3D platform, the virtual twin model of the working face and equipment with the same size of the corresponding entity, the kinematics model of the cutting unit and the virtual collision detection model are established. The virtual model movement is controlled through the intelligent interactive interface at the virtual end, and the teaching trajectory is designed and optimized according to the worker’s experience, so that it can be used as the target expected trajectory of trajectory tracking to make up for the excessive dependence on the worker’s experience caused by the traditional underground manual teaching due to the harsh working conditions. Secondly, in order to improve the quality of section forming, the control method of teaching trajectory tracking and reproduction in memory automatic cutting stage is studied. The dynamic model of cutting part is established by Lagrange method, and the tracking control accuracy of end effector to teaching trajectory is improved by combining iterative learning with sliding mode control. Finally, the simulation control platform of the memory cutting of the cantilever roadheader is built. Through the real-time data transmission and interaction between the virtual space and the physical space and between the modules, the three-dimensional visual simulation of the memory cutting virtual teaching and trajectory tracking control process is completed in the virtual space, and then the memory automatic cutting trajectory tracking control command is generated and sent to the end effector of the physical entity of the cantilever roadheader to drive it to carry out the section forming cutting according to the teaching trajectory. At the same time, the physical sensor collects the pose data of the cantilever roadheader fuselage and the cutting arm, and reversely drives the virtual model to move synchronously. The closed-loop control of robot virtual model and physical entity is realized. On this basis, the virtual and real synchronization of the system, the motion consistency between the virtual prototype and the physical prototype, and the trajectory tracking and reproduction control accuracy are verified. The experimental results show that the system data transmission delay is low, which can ensure the virtual and real consistency and synchronization, and the trajectory tracking control accuracy meets the actual use requirements. This method provides a new idea for memory cutting and intelligent control of tunneling equipment.

国家自然科学基金青年基金资助项目(52104166)；国家自然科学基金面上资助项目(52174105)；陕煤联合基金资助项目(2021JLM-03)

张旭辉，王甜，张超，等. 数字孪生驱动的悬臂式掘进机虚拟示教记忆截割方法[J]. 煤炭学报，2023，48(11)：4247−4260.

ZHANG Xuhui，WANG Tian，ZHANG Chao，et al. DT-driven memory cutting control method using VR instruction of boom-type roadheader[J]. Journal of China Coal Society，2023，48(11)：4247−4260.