Study On Prediction And Early-warning Of Multi-source Information Of Rock Burst In Hard Roof Of Tang Kou Coal Mine

Coal is the main energy source for China's economic construction.With the continuous increase in mining intensity and scale over the years,China's shallow coal resources have been reduced year by year,and the exploitation of deep coal resources is the trend of future development.However,deep mining has serious challenges in disaster prevention and control.One of these is the rock burst geodynamic disaster.This dissertation took the induced impact and prevention of hard roof working face in Tangkou Coal Mine as the research background.The theoretical analysis,numerical simulation and on-site monitoring data analysis methods have used to analyze the mechanism of inducing rock burst hazard of hard roof and study the bending,breaking and movement of hard roof.This dissertation established a prediction and early warning model for rockburst monitoring,which provided a theoretical basis and guiding significance for the safe mining of the hard roof working face of mines.The main research is as follows:(1)According to energy theory analysis,the critical energy density required for impact rock pressure in the hard roof working face of Tangkou Coal Mine has obtained;The mechanical model of impact instability of hard roof has analyzed;the maximum stress concentration area and bending moment position in front of the coal wall of the 5305 hard roof working face has calculated.(2)This dissertation has establish a simulation model of hard roof working face by numerical simulation,study bending-breaking-movement law of hard roof during mining process and distribution of advance bearing pressure in front of working face coal wall,elastic strain energy formation accumulation and release evolution rules provide guidance for the safe mining of working face.(3)Through on-the-spot collection and analysis of the rock burst monitoring data of 5305 working face in Tangkou Coal Mine,and based on wavelet neural network theory,a prediction model of rock burst has established.based on the principle of clustering algorithm,and the microseism Energy,drill cuttings,and stress values were used for cluster prediction analysis.Through the establishment of impact rock pressure monitoring data prediction and mining analysis models,the mine rock pressure monitoring data was cilcctively and reasonably used.(4)Based on the theory of fuzzy neural network,the multi-source information feature extraction and analysis of the rock burst monitoring data has used to establish a fuzzy neural network comprehensive monitoring and carly warning model for multi-source information of rock burst,and the early warning model has applied in 5305 working face.After on-site inspection and analysis,the fuzzy neural network early warning model works well and the accuracy of carly warning was high.