Performance prediction of tunnel boring machine through developing high accuracy equations: A case study in adverse geological condition

摘要

The aim of present study is to propose new superior equations and introduce novel techniques for TBM performance prediction. To this end, correlations between the Rate of Penetration (ROP) and rock mass properties are investigated using four simple regression analyses. Based on these analyses, two nonlinear multivariable equations are introduced and optimized by the Imperialist Competitive Algorithm (ICA). Then, two other distinct models are examined by using the Classification and Regression Tree (CART) and Genetic Expression Programming (GEP) techniques. The aforementioned methods are applied on a dataset from the Queens Tunnel, in New-York City with complex geology conditions. It was found that the models proposed by ICA, CART and GEP techniques have determination coefficient of 0.76, 0.82 and 0.72 for training data, and 0.62, 0.69 and 0.65 for testing data, respectively. The results showed the noticeable improvement of the predictions compare to previous studies. (C) 2019 Elsevier Ltd. All rights reserved.