Estimating maximum surface settlement due to EPBM tunnelling by Numerical- Intelligent approach- A case study: Tehran subway line 7

نویسندگانSayed Rahim Moeinossadat- Kaveh Ahangari
نشریهTransportation Geotechnics
ارائه به نام دانشگاهDepartment of Mining Engineering, Science and Research Branch, Islamic Azad University, Tehran, Iran
شماره صفحات92-102
شماره سریال18
نوع مقالهFull Paper
تاریخ انتشار2019-03-01
رتبه نشریهISI (WOS)
نوع نشریهچاپی
کشور محل چاپهلند
چکیده مقاله<p dir="ltr" id="sp0005" style="text-align: justify;">Ground&nbsp;<a href="https://www.sciencedirect.com/topics/social-sciences/settlement" title="Learn more about Settlement">settlement</a>&nbsp;due to&nbsp;<a href="https://www.sciencedirect.com/topics/engineering/excavation" title="Learn more about Excavation">excavation</a>&nbsp;of&nbsp;<a href="https://www.sciencedirect.com/topics/engineering/shallower" title="Learn more about Shallower">shallow</a>&nbsp;tunnels is a common phenomenon. To control the settlement, one should be able to predict it, and based on it he may consider required preventions and protections. There are different methods for predicting settlement, each having some strengths and weaknesses. The main weakness of these methods is that they do not consider enough effective parameters on the settlement. The&nbsp;<a href="https://www.sciencedirect.com/topics/social-sciences/numerical-analysis" title="Learn more about Numerical Analysis">numerical methods</a>, contrary to the empirical and analytical methods, take into account the effects of a larger number of parameters. However, ideal selection of many parameters is associated with ambiguity and difficulty and is time-consuming. To overcome these issues, the intelligent methods are incorporated which are appropriate tools. The aim of this paper is to present a numerical-intelligent model for prediction of maximum surface settlement (<em>S<sub>max</sub></em>). At first, a section of Tehran&nbsp;<a href="https://www.sciencedirect.com/topics/engineering/subways" title="Learn more about Subways">subway</a>&nbsp;line 7 was modeled using the&nbsp;<a href="https://www.sciencedirect.com/topics/engineering/finite-difference-method" title="Learn more about finite difference method">finite difference method</a>&nbsp;(FDM). Then a&nbsp;<a href="https://www.sciencedirect.com/topics/engineering/dataset" title="Learn more about Dataset">dataset</a>&nbsp;including 100&nbsp;<em>S<sub>max</sub></em>&nbsp;values were prepared for creating the intelligent model. Among the intelligent methods, the gene expression programming (GEP) method was selected to represent the&nbsp;<a href="https://www.sciencedirect.com/topics/engineering/mathematical-equation" title="Learn more about Mathematical Equation">mathematical equation</a>and the built numerical-intelligent model explained a proper performance. The determination coefficient, R<sup>2</sup>, for both the training and testing phases was 0.976 and 0.931, respectively. At the end, the derived mathematical equation from the GEP model was prepared using the&nbsp;<a href="https://www.sciencedirect.com/topics/engineering/visual-basic" title="Learn more about Visual Basic">visual basic</a>&nbsp;(VB) in the form of predictor software. According to accuracy of the prediction results, the presented equation and software are reliable and suitable as an alternative for the&nbsp;<a href="https://www.sciencedirect.com/topics/engineering/numerical-modelling" title="Learn more about Numerical Modelling">numerical modelling</a>.</p> <p>&nbsp;</p> <ul dir="ltr" id="issue-navigation"> </ul>

لینک ثابت مقاله

tags: Subway tunnel; Surface settlement; FDM; GEP; VB