Search

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.