A Research On The Prediction Model Of Mine Safety Accident Based On The Theory Of Gray System

With China’s rapid development of industry, the demand of the mineral resourcesis increasing. In the mine production process, all kinds of accidents happen frequentlyserious threat to mining personnel life safety, production situation is not optimistic,affects the mineral industry development. In view of the mine accident hasuncertainty, this paper uses gray system theory and statistical knowledge, this paperemphatically analyzes and construct the basic mine accident prediction models for thecoal mine production safety, will provide a certain reference value. The main contentsare as follows:(1) The mine accidents and injury rate is the main form of safe production, inorder to study the mine production safety situation of the work carried out smoothly,this thesis on mine production safety accidents over the years of data collection andstatistics.(2) Casualty rate is the embodiment of mine production safety situation ofcomprehensive index, gas emission characterization of mining in a certain period oftime interval production safety coefficient. In this paper, the theory of grey systemgrey GM (1,1) model is applied to a mine accident casualty rate and the prediction ofthe gas emission quantity, between traditional prediction model deficienciesintroduced a Markova process of Markov chain theory and method of buffer operatorto improve the model, further improve the prediction results of the precision andaccuracy, and through this paper analysis.(3) Based on the grey system theory, accident prediction models, design of coalmine safety accidents grey prediction algorithm and prediction model for work flow,for the future of the coal mine safety production management work to provide aneffective auxiliary platform.In this paper on the basis of grey model has been further developed, theimproved model is applied to a specific mine to forecast, the results show that, using amodified combination forecasting model, the relative errors are below8%, obtain agood forecast effect; and the single grey methods improved grey model, the ability toadapt and promote better prediction ability, higher prediction accuracy.