Abstract: Aiming at the existing problems of relatively large errors,great influence by the environment,relatively large adjustment deviations,unstable results and others in the traditional positioning and detection method of roadheader postures in Yongdingzhuang Coal Industry Co.,Ltd.,a study on positioning and adjustment methods of roadheader body postures is conducted.Numerical analysis method is used to analyze the rigidity postures of roadheaders and coordinate transformation is carried out.Machine vision technology is used to design a precise positioning and automatic adjustment system for roadheader postures.The roadheader posture visual signals collected on site are intelligently guided to control the execution mechanism of the roadheader,achieving roadheader precise positioning and posture automatic adjustment.Hardware system type selection and software control system design are completed,and the roadheader’s own motion trajectories and set values are autonomously controlled.After conducting on-site application in the downhole of Yongdingzhuang Coal Industry Co.,Ltd.,it shows that the cutting head of the roadheader posture automatic positioning system adopting robot technology and machine vision technology can automatically track the target trajectories,adjust the posture state of the roadheader according to the set trajectories,and autonomous optimize the walking paths of the roadheader.After repeated experiments,the error between the obtained experimental results and simulation results is less than 4.56 mm,achieving satisfactory experimental results and meeting the high-precision positioning requirements of on-site roadheaders in coal mine downholes.