Abstract: Intelligent perception of coal caving volume is the key technology for intelligent fully-mechanized caving mining. Through the real-time monitoring of coal flow of the scraper conveyor at the rear of the fully-mechanized caving face, the action of the coal caving port can be dynamically regulated. Combined with coal gangue identification and top coal thickness monitoring information, over-discharge and under-discharge of the working face could be avoided effectively. By using the proposed method, it is easy to improve the efficiency of coal resource mining and recovery, and prevent the occurrence of safety production accidents such as overloading of scraper conveyors. In this paper, an intelligent monitoring method of coal caving volume in fully-mechanized caving face based on laser scanning was proposed, and the triangular micro-element method was introduced to construct a real-time calculation model of coal caving volume regression. To characterize the real-time coal discharge at the working face, high-performance multiple echo signal reflection lidar scanning was used to quickly capture and store high-precision 3D laser point cloud data, and a regression processing algorithm based on the least squares method for laser echo data was proposed. The control experiment of the spline interpolation algorithm shows that the calculation accuracy and efficiency of the data regression algorithm have obvious advantages. At the same time, the robustness of the algorithm was verified in the laboratory environment. The 8222 working face of Tashan Coal Mine of Jinneng Group has carried out an industrial test of the intelligent monitoring system of coal discharge, which has verified the reliability of the monitoring technology and device.