Abstract: Rock burst situation assessment is one of the key and difficult issues in the safety monitoring and early warning field of coal mining. In this paper, based on real-time data from microseism, acoustic emission and electromagnetic radiation monitoring, situation assessment theory was introduced to the rock burst prediction, and a new situation assessment model of rock burst was established based on log audit of monitoring system and real-time correction algorithm. In addition, the situation quantitative method of rock burst was put forward by improving D-S evidence theory. Moreover, by using this method to calculate situation values in continual time points, the predictive values of rock burst situation were calculated and analyzed based on GM (1, 1) and ARMA models, respectively. Test results and related analysis show that the proposed situation prediction method of rock burst is reasonable and effective, and can reflect the development trend of rock burst dynamically.

Key words: rock burst, information fusion, situation assessment, D-S evidence theory, log audit