Abstract: Aiming at the problems of traditional memory cutting overbreak and underbreak, low resource recovery rate, and roof management risks caused by local roof and floor undulations, microstructure development in the 2-meter medium-thick coal seam working face of Gansu Lingtai Shaozhai Coal Industry Co., Ltd, based on the multi-source information fusion theory and predictive control method, an intelligent cutting technology integrating autonomous planning and advanced correction is developed and applied. This technology by integrating fiber optic inertial navigation and multi-sensor array, and adopting extended Kalman filtering algorithm to achieve high-precision real-time perception of shearer poses; Use Kriging interpolation algorithm to construct and dynamically refresh a 3D geological model of the working face; Apply the improved A* path finding algorithm to autonomously plan the next optimal cutting curve based on dynamic model; Establish a quantitative correction model that integrates realistic information from downhole engineering to achieve advanced intervention for sudden geological anomalies. The industrial tests show that this technology reduces the root mean square error of cutting height from 142.6 mm to 37.5 mm, improves the resource recovery rate of the working face from 95.1% to 97.6%, the gangue content of raw coal is decreased by 61.1%, and the daily average advancement progress is improved by 24.6%, which effectively solve the difficult problem of precise cutting in intelligent mining of this mine and provide reliable technical support for safe, efficient, and green mining of medium-thick coal seams.