陕北空洞型采空区稳定性评价——以讨老乌素煤矿采空区为例.pdfVIP

ofcavitary gob area were choosen as inputneurons ofthemodel ofBPneural network and the results of stability of gob area were choosen as output neurons. Then the number of middle neurons were defined by trial computing with the golden section method and the empirical method. At last, the structure of BP neural network model was defined as 11—22—4, The model which had the best ability to distinguish was made by training and amending with samples of cavitary gob area which were collected. Then the model was used to evaluate the stabilityofTaolaowusucoalgob areaandtheresultwas “instability”. The roof and floor of coal seam model of Taolaowusu coal gob area were built through the Finite difference software FLAC3D, then the process of excavating coal seam and loading roadbed when the gob area had not managed were simulated under the initial stress field and the displacement, stress and plastic developments of model s were observed.According to the theory of rock strength, the results which the roof and floor of coal seam is safe and the coal block is in danger when the compressive stress of coal block tended to be close to strengthen of coal blcok. Especially in that case which the gob area had not managed，the coal block would be destoryed instantaneously when loading roadbed on the floor. Therefore, the evaluated resultofnumericalexperimentationwas instability too. “ ” Keywords: Keywords: KKeeyywwoorrddss::CavitaryGobArea Factor BPneural Network Numerical ExperimentationEvaluation Thesis : Thesis : TThheessiiss :: ApplicationResearch 1绪论 1绪论 1.1 选题背景及研究意义 1.1.1 选题背景 陕西北部包括延安与榆林两市，地处黄土高原，矿产资源丰富。以优质煤、石油、 天然气、石灰岩、粘土类及盐类矿产为主。而其中煤炭资源是其重中之重，优中之优。 陕北煤炭资源主要是以榆林为中心，分为：神府、榆横、榆神三大矿区。该煤炭在神木、 府谷、横山县和榆阳区境内埋藏浅，在靖、定边两县境内埋藏较深（地表500m以下），