A new method for forecasting shield's disc-cutters wearing based on Elman neural network

作者：Xiaojiang Su · 发表于：Journal of Liaoning Technical University · 年份：2010 · 被引用次数：6 · 研究领域：Image Processing and 3D Reconstruction、Tunneling and Rock Mechanics、Industrial Technology and Control Systems

To overcome the difficulty in estimating the wearing of shield's disc-cutters in composite geological ground conditions,a new method for forecasting the wearing of disc-cutters using Elman neural network model is proposed in this paper.The model is built using the shield excavation parameters data at the stage of normal wearing of newly replaced disc-cutters.The disc-cutters wearing is forecasted according to the difference between the predicted excavation speed from the model and the actual excavation speed.In this paper,the influencing factors of disc-cutters wearing are analyzed,the structure of Elman neural network model is determined and the criterion of estimating disc-cutters wearing is presented.A case study was conducted in the shield construction project at Caotao left-line of Guangzhou Metro Line 5.The study results prove that the forecast result is in line with the actual situation of disc-cutters replacment.The model is simple,effective and feasible.Also it has strong adaptive ability.This study can provide a useful guidance for estimating disc-cutters wearing and replacing under similar geological conditions.