The Reseach On Neural Network Predict The Laneway Distortion

In this thesis, the author made a system analysis of the distortion of the laneways,and the author made predictions to the distortion of the laneway by three different neural networks.We have evolved the upsides and downsides of three different neural networks, finally the use of GRNN to make quantitative predictions of the distortion of the laneway in Fengfeng Coal Mine Area, and the result of the experiment is in accordance with the theoretical prediction. The predict consequences provided the basis for construction and production of the mine.Drring the period of the thesis, the author did lots of work to gather data, observe in the port, experiment, reference lots of documents and data, and gained the following achievements.(1) The author made in-depth and careful studies and analysis on the mechanism of the soft rock in laneways, attending the study of supporting technology in soft rock tunnel with complex stress and high rheodynamics of joints. Then the author questioned the method for the distortion of the laneway.(2) The author analyzed the factor which affected the distortion of the laneways, then there should be a basic summary of our past work so that we intends to use the factor which affected the distortion of the laneways in research of prediction.(3) Though the study to neural networks, understanding the mathematical of neural networks, analysis upsides and downsides of a variety of different algorithms of neural networks, finally we have worked out best ways of predicting of the distortion of the laneway.(4) There are many factors affected the distortion. Some of them we know a lot, but some of we know little. And the factors affect each other. So we need fuzzy algorithm to solve this problem. The neural network algorithm can predict the distortion because of its powerful capability of non-linearity and dunamic disposal. From the calculated result, we know the algorithm is suit to solve the problem.(5) First and last we adapted to three neural networks. Using BP ,RBF and GRNN neural networks to predict the laneway distortion, and we have worked out best ways of predicting of the distortion of the laneway. (6) Attending the study of combination of anchorage and shotcrete support in soft rock tunnel with complex stress and high rheodynamics of joints, the author list the posiible factors affecting the laneway distortion. Then we used some important factors to predict the distortion.(7) In general, analyzing the factors of laneway distortion, using GRNN neural nerwork algorithm to calculate, predicting the value of laneway distortion using GRNN neural network can gain a better result.