Abstract: In view of the scientific problems and key technologies that restrict the advance and accurate early warning of coal mine dynamic disasters,such as coal mine rock burst,coal and gas outburst,water inrush and water burst in coal mines,this paper proposes the original source early warning method of coal mine dynamic disasters and the information system realization technology. The basic idea is that on the basis of accurate geological modeling and real-time dynamic update of mining space,the macroscopic qualitative description of the mechanism and evolution law of coal mine dynamic disasters and the related generalized models are transformed into online computable mathematical and mechanical expression model through mathematical converter. The key technologies for achieving an accurate early warning of coal mine dynamic disasters are as follows:①Through the research and development of hologram sensing technology,the optimal deployment and installation of sensors can accurately collect information such as microseism,absolute geo-stress,mine pressure,displacement,electromagnetic radiation,gas emission quantity,gas extraction quantity,water inflow,water level,water pressure and water quality,etc.,and the precursory presentative information of coal mine dynamic disasters is monitored in real time. ②By means of the research and development of automatic recognizer for complex information,the model and algorithm of processing monitoring data,such as data decomposition,data filtering,data enhancement,data identification,data interpolation,data anti-drift and data reconstruction,can be realized,and all kinds of collected presentation information are identified,so as to achieve the purpose of eliminating the false and retaining the true,taking the essence and discarding the dregs. ③By researching the precision converter,the transformation from the presentation information,such as microseism,absolute geo-stress,mine pressure,displacement,electromagnetic radiation,gas emission quantity,gas extraction quantity,water inflow,water level,water pressure and water quality,to the original information closely related to the coal mine dynamic disaster,such as surrounding rock stress and potential distribution,geometric deformation of mining disturbance,gas content and gas pressure distribution,water-rich quantity and water-rich pressure distribution,porosity of coal seam and surrounding rock,compressive strength,tensile strength and elasticity modulus,can be realized. ④Through the research and development of intelligent and accurate early warning devices for multiple complex dynamic disasters in coal mines,making full use of dynamic original source information,the advanced and precise prediction and early warning of dynamic disaster in coal mines such as rock burst,coal and gas outburst,water inrush and water burst can be realized. ⑤By establishing a precise early warning platform for major dynamic disaster in coal mines,using big data,cloud computing and artificial intelligence technology,a precise early warning service platform for major dynamic disasters in coal mines based on SOA framework is established,which can realize the access,processing,prediction and early warning and push service of the precursory presentative information of coal mine dynamic disasters in mines,group companies and even the whole country. The main purpose of this paper is to overcome the "technical bottleneck" and tough problems of accurate prediction and early warning of major dynamic disaster in coal mines,integrate existing research and development achievements,and develop an accurate early warning platform for major dynamic disasters in coal mines to achieve an advanced perception,precise early warning and prediction of major dynamic disasters in coal mines.