Research on prediction of water conducted fissure height in roof of coal mining seam

马亚杰,武强,章之燕,洪益清,郭立稳,田洪胜,张丽阁
DOI: https://doi.org/10.13199/j.cst.2008.05.70.mayj.022
2008-01-01
Abstract:In order scientifically to design the size of the water proof coal (or rock) pillars, the paper discussed and researched the prediction method of the water conducted fissure zone height in the coal mining face roof. According to the difficult quantitative and complicated non-linear relationship between the water conducted fissure zone height in the roof and each influence factor, a BP artificial nervous network method with non linear shining, extensive and error tolerance functions was adopted. 8 predicted indexes collected with the measured data formed sample collection base on the roof water conducted fissure zone height in Kailuan and other 18 mining areas were applied to establish a three layers network to predict the water conducted fissure zone height. The network was passed the verification with a minimum error. Each index influenced weighting distribution obtained from the network calculation were well met the present research understanding. Therefore, the method is rational and suitable.
What problem does this paper attempt to address?