REZA MIKAEIL - Faculty of mining , Petroleum & Geophysics, Shahrood University of tech, Daneshgah Blvd, shahrood, Iran
Omid Frough - Faculty of mining , Petroleum & Geophysics, Shahrood University of tech, Daneshgah Blvd, shahrood, Iran
Reza Khalokakaie - Faculty of mining , Petroleum & Geophysics, Shahrood University of tech, Daneshgah Blvd, shahrood, Iran
Mohammad Ataei - Faculty of mining , Petroleum & Geophysics, Shahrood University of tech, Daneshgah Blvd, shahrood, Iran

The prediction of performance of tunnel boring machines (TMB) penetration rate is important for project planning and selection of economic tunneling methods.This paper presents an attempt to predict penetration rate of TBM with a generalized regression neural network. The Queens Water tunnel data have been used to develop this network which includes three layers (input, hidden and output layers). The compressive trength, peak slope index, distance between planes of weakness and orientation of discontinuities in rock mass are chosen as input data penetration rata of TBM as output data. The results show that develop network is capable of predicting TBM penetration rata with correlation coefficient of ۰.۹۱۱. It was concluded that the penetration rata can be reliably estimated using the generalized neural network.

Tunnel Boring Machine ; Penetration Rata ; Generalized Regression Network ; Queens Water Tunnel