A Predictive Model for Estimating the TBM Penetration Rate Based on Hybrid ICA-ANN and DEA-AHP Algorithms

摘要

Predicting the penetration rate of tunnel boring machine (TBM) is a complex and challenging task that plays a crucial role in the schedule planning and cost estimation of tunneling projects. In this study, Mutual Information (MI) is utilized to select the most effective parameters on penetration rate among various rock mass and machine operating parameters. Moreover, Data Envelopment Analysis-Analytic Hierarchy Process (DEA-AHP) approach is implemented in Imperialist Competitive Algorithm-Artificial Neural Network (ICA-ANN) to improve its performance as well as to achieve the best possible architecture of the network. The aforementioned methods are applied on a database consists of 430 data collected from the Lot 2 of Zagros tunnel project. The results indicated that the network with 6 input variables, including uniaxial compressive strength (UCS), Brazilian tensile strength (BTS), distance between the planes of weakness (DPW), rock quality designation (RQD), thrust and revolution per minute (RPM), 9 neurons in the hidden layer and the penetration rate as the output provides the best performance. The determination coefficient (R-2) and the mean square error (MSE) indices of the training and testing datasets are 0.901, 0.887, 0.04 and 0.0645, respectively. This study shows that the aforementioned methods could be utilized for enhancing the prediction of the penetration rate of TBMs.