Abstract: Aiming at the difficult positioning and difficult orientation of the tunneling machine in the narrow and narrow closed roadway of coal mine,the laser radar-based underground roadway environment modeling and obstacle detection method based on laser radar is studied to judge the shape of the roadway edge and predict the direction of the roadheader.The kinematics model and lidar observation model of coal mine roadheader are established,and the excavation distance and position of the tunneling machine in the coal mine roadway are calculated.The modeling problem of narrow and closed coal mine roadway environment is transformed into mathematical probability and statistics.Predict problems with mathematical models.At the same time,the laser SLAM environment modeling algorithm is analyzed,and the feasible roadway environment modeling algorithm is optimized and improved to further improve the accuracy of line feature extraction in underground roadway environment modeling.The experimental results show that the laser radar-based coal mine roadheader environment modeling and obstacle detection method can effectively solve the problem of no map guidance in the tunneling process of the tunneling machine.When the distance_thresh is 0.05 and the angle_thresh is 0.06,the HectorSLAM algorithm can accurately extract the piece.Help the shape of the edge of the roadway,identify the obstacles in the coal mine roadway,and provide guarantee for predicting the direction of the roadheader and judging the shape of the roadway.