Multi-sensor Fusion Positioning System and Experimental Study of Roadheader

The autonomous and accurate positioning of roadheader is the foundation of intelligent development of coal mine roadway excavation. However，the complicated tunneling technology and harsh tunneling environment make the positioning of roadheader faces problems such as insufficient autonomy，low accuracy and poor immunity to disturbance. In order to realize the autonomous and accurate positioning of roadheader，an odometer-aided inertial positioning system is constructed based on extended Kalman filter to restrain the error divergence of pure inertial positioning. Combined with the construction technology of roadheader，a flexible zero velocity updating method is proposed to further improve the inertial positioning accuracy of roadheader. The efficient integration of inertia，vision and mileage positioning is realized based on error state Kalman filter and multi-state constraint model. A prototype test system of roadheader in dark environment is also built. The experimental results show that the proposed flexible zero velocity updating method can improve the pure inertial positioning accuracy by 21.64%. The three-axis positioning errors of the proposed multi-sensor fusion positioning system of roadheader are within 0.13 meters of lateral positioning error，0.17 meters of forward positioning error and 0.02 meters of upwards positioning error，which improves by 49.62% and 57.71% compared with the independent inertial and visual systems，respectively. The experimental results verify the feasibility and effectiveness of the proposed method and system，and the positioning accuracy of the proposed system meets the requirements of roadway excavation in coal mine. © 2023 Nanjing University of Aeronautics an Astronautics. All rights reserved.