The Research Of Information Fusion Method Technology In Coal Mine Safety Monitoring

Coal resources is always the most important resources during the past few decadesin our coutry. However, we also have the highest incidence of coal mine accidents in theworld.These disasters not only bring tremendous losses to property security but also topeople life security. So, it is deserved to study on using information technology to helpthe Coal Safety Production. Nowdays,the main coal mine accidents consist of gasexplosion、dust explosion、fire disaster and roof caving accidents. And the kinds ofexplosion accidents are the hardest disasters.In the coal mine roadway, there are many kinds of sensors, they are real timemonitoring each index of the coal mine. But, the conditions in the mine is very bad,single sensor may lose efficacy and make erroneous judgement. And, there arenonlinear relationships between different kinds of sensors. So, we use informationfusion technology to improve the prediction efficiency of coal mine environment.This paper presents an information fusion model with two levels of structure. Weuse it to predict the Coal mine safety environment.This model three kinds of data fusionincluded gas information、dust information and wind speed information. At the firstlevel of this model, we use weighted adaptive algorithm to calculate the differentweights between the same sensors. The algorithm could improve the accuracyexpression of these sensors. At the second level of this model, we use Fuzzy NeuralNetwork to complete the predict function.In the one hand, the algorithm can deal withthe fuzzy information, in the other hand, it has a learning disability, and it could changewith the environment and updated. On the whole, the information fusion model couldmeet the actual environment of the coal mine, and it could effectively improve theprediction accuracy.In order to improve the operation speed of the information fusion model, the paperdeeply study on a kind of improved LMBP algorithm. Through many experiments, thepaper detailed analysis the main existing problems of the traditional LMBP algorithm,and an improved algorithm is put forward through introducing two distinguishingfactors.Through the Matlab simulation under the condition of experiment,it is beingproved that the improved LMBP algorithm has relatively substantially improved inthe speed.In addition, by the model prediction experiment, it shows, the predictionaccuracy of the prediction model could be more than96%. At last, in this paper, a kind of safety monitoring system is designed by usingASP.net technology.This system could real time monitor the gas information、dustinformation、wind speed information and so on. And it could analysis of the data andprint these information by using report forms and chart. This system is hybridprogrammed by using C#and MATALB language.