On the pointlessness of machine learning based time delayed prediction of TBM operational data

Abstract In tunneling, predictions of the rockmass conditions ahead of the face are of great interest to be able to take appropriate countermeasures at the right time. Besides investigations like drilling or geophysics, new approaches in mechanized tunneling aim at forecasting the geology ahead via Machine Learning models. These models are trained to forecast tunnel boring machine data by learning from recorded data in already excavated parts of the tunnel. Simply judging from high accuracies, these results may look promising at the first sight, but forecasts like this are mostly just delayed and slightly altered versions of the input data and no predictive value can result from them. This paper shows deficits in the current practice of data driven forecasts ahead of the tunnel face and raises impetus for further research in this particular field and TBM data analysis in general.

[1]  SchmidhuberJürgen,et al.  2005 Special Issue , 2005 .

[2]  V. Panchang,et al.  One-Day Wave Forecasts Based on Artificial Neural Networks , 2006 .

[3]  Alexander Radinger,et al.  Consistent exploration by probe drilling and TSWD through the example of the Koralm Tunnel / Konsequente Vorauserkundung mittels Bohrungen und TSWD am Beispiel des Koralmtunnels , 2014 .

[4]  George Athanasopoulos,et al.  Forecasting: principles and practice , 2013 .

[5]  K. V. Uday,et al.  Predictions of Weekly Soil Movements Using Moving-Average and Support-Vector Methods: A Case-Study in Chamoli, India , 2019, Information Technology in Geo-Engineering.

[6]  Konrad Bergmeister,et al.  Learning and optimization from the exploratory tunnel – Brenner Base Tunnel , 2017 .

[7]  Leo Breiman,et al.  Random Forests , 2001, Machine Learning.

[8]  Rob J Hyndman,et al.  Another look at measures of forecast accuracy , 2006 .

[9]  Thomas Marcher,et al.  Application of artificial neural networks for Underground construction – Chances and challenges – Insights from the BBT exploratory tunnel Ahrental Pfons , 2019, Geomechanics and Tunnelling.

[10]  Richard Hans Robert Hahnloser,et al.  Digital selection and analogue amplification coexist in a cortex-inspired silicon circuit , 2000, Nature.

[11]  Jürgen Schmidhuber,et al.  Long Short-Term Memory , 1997, Neural Computation.

[12]  Pradnya Dixit,et al.  Removing prediction lag in wave height forecasting using Neuro - Wavelet modeling technique , 2015 .

[13]  Kuldip K. Paliwal,et al.  Bidirectional recurrent neural networks , 1997, IEEE Trans. Signal Process..

[14]  Michael Y. Hu,et al.  Effect of data standardization on neural network training , 1996 .

[15]  Stefan Lessmann,et al.  A comparative study of LSTM neural networks in forecasting day-ahead global horizontal irradiance with satellite data , 2018 .

[16]  Samuel T. Ariaratnam,et al.  Applying radial basis function neural networks to estimate next-cycle production rates in tunnelling construction , 2010 .

[17]  Gaël Varoquaux,et al.  Scikit-learn: Machine Learning in Python , 2011, J. Mach. Learn. Res..

[18]  Vladimir N. Vapnik,et al.  The Nature of Statistical Learning Theory , 2000, Statistics for Engineering and Information Science.

[19]  Geoffrey E. Hinton,et al.  Deep Learning , 2015, Nature.

[20]  Jinjun Tang,et al.  Evaluation of Short-Term Freeway Speed Prediction Based on Periodic Analysis Using Statistical Models and Machine Learning Models , 2020 .

[21]  Nick Barton,et al.  Engineering classification of rock masses for the design of tunnel support , 1974 .

[22]  Mahmoud Hassanlourad,et al.  Evaluation of liquefaction potential based on CPT data using random forest , 2015, Natural Hazards.

[23]  Konrad Bergmeister,et al.  Development of holistic prognosis models using exploration techniques and seismic prediction , 2017 .

[24]  Chao Zhang,et al.  Recurrent neural networks for real-time prediction of TBM operating parameters , 2019, Automation in Construction.

[25]  Han Cao,et al.  Prediction for Tourism Flow based on LSTM Neural Network , 2017, IIKI.

[26]  Francisco Alonso-Sarría,et al.  Modification of the random forest algorithm to avoid statistical dependence problems when classifying remote sensing imagery , 2017, Comput. Geosci..

[27]  M. J. Fuente,et al.  Monitor-While-Drilling-based estimation of rock mass rating with computational intelligence: The case of tunnel excavation front , 2018, Automation in Construction.

[28]  T. Rientjes,et al.  Constraints of artificial neural networks for rainfall-runoff modelling: trade-offs in hydrological state representation and model evaluation , 2005 .

[29]  R. Calvo,et al.  Neural Network Prediction of Solar Activity , 1995 .

[30]  Kenneth E. Williams,et al.  Prediction of solar activity with a neural network and its effect on orbit prediction , 1991 .

[31]  Jitendra Kumar,et al.  Long Short Term Memory Recurrent Neural Network (LSTM-RNN) Based Workload Forecasting Model For Cloud Datacenters , 2018 .

[32]  Dietmar Bach,et al.  The use of TBM process data as a normative basis of the contractual advance classification for TBM advances in hard rock , 2018 .

[33]  Bernd Moritz,et al.  Determination of the system behaviour based on data analysis of a hard rock shield TBM / Analyse der Maschinenparameter zur Erfassung des Systemverhaltens beim Hartgesteins‐Schildvortrieb , 2014 .

[34]  Kourosh Shahriar,et al.  A support vector regression model for predicting tunnel boring machine penetration rates , 2014 .

[35]  John C. Brown,et al.  Delayed time series predictions with neural networks , 1998, Neurocomputing.

[36]  Thomas Marcher,et al.  Artificial Neural Network Based Online Rockmass Behavior Classification of TBM Data , 2019, Information Technology in Geo-Engineering.

[37]  Yen-Ming Chiang,et al.  A two‐step‐ahead recurrent neural network for stream‐flow forecasting , 2004 .

[38]  Jürgen Schmidhuber,et al.  Framewise phoneme classification with bidirectional LSTM and other neural network architectures , 2005, Neural Networks.