Development And Research Of Discrimination And Prediction System For Deep Seam Mining Floor Water-Bursting

With the gradual development of deep coal, the hazard of deep seam mining floor water-bursting is more sigificiant in a complex hydrogeological environment and to mine the deep coal seam safely is much more difficult then before. In this paper, the integrated approaches of factorial analysis, mathematical model and system program were adopted to study the main factors, discriminant model and prediction system.The critical depth of deep mine was calculated and each factorial function of water-bursting was analysed, then assessment index system was derived.By AHP(analytic hierarchy process), it got the main factors of deep-well floor water bursting. The AHP automatical adjusting model program was built to solve the question that the dynamical judgement matrix's globe consistece is usually hard to satisfy. Every factor's information data were normalized by the membership function or membership grade, the FNN (fuzzy neural network) distinguishment model was taken out. Choosing the proper network's paramater was used to avoiding neural network its weakness, its reliability was tested by the engineering projects.The chosen main factors were verified by the grey relevance analysis, the contribution weights was ascertained by the entropy adjusting the ancient. According to professor Xiao Fangchun's grey matter-element analysis theory, the linear grey matter-element model and optimal grey matter-element model were established. By the means of the engineering projects, the thresholds of safe zone, vulnerable zone and dangerous zone were ascertained.Based on the linear grey matter-element and optimal grey matter-element model, the discrimination system for deep seam mining floor water-bursting combined grey matter-element is exploited by VB 6.0. It can be divided into the four parts:information management, water-bursting judgement, system maintenance and system service. Its function was comparely perfect and has character of simplicity, practicality and reliability, the prediction result was much comparely scientific and objective.