To reduce the probability of geological risks during tunneling work using the shield TBM, prospecting methods have been proposed to predict ground conditions ahead of a tunnel face. However, these methods that have been proposed so far usually require additional time for testings and measurements performed when the TBM driving is not in operation. Therefore, prospecting is usually limited to suspected areas of risky ground identified in the design stage. In order to overcome this drawback, this study presents a ground condition prediction technique ahead of the tunnel face that only requires the operational data of the shield TBM and avoids the need for additional testing or measurement. For this purpose, the following topics are studied in this thesis:
Firstly, an artificial neural network (ANN) was used to analyze the operational data of the shield TBM acquired during tunnel excavation stage to predict ground conditions ahead of a tunnel face regardless of site conditions. The primary advantage of the proposed technique is that, by using TBM data, no additional data acquisition device is required. Ground type classifications and machine data normalization methods are introduced to maintain the consistency of the measured data and improve prediction accuracy. The efficacy of the proposed model is demonstrated by its 96% accuracy in predicting ground type one ring ahead of the tunnel face.
Secondly, time series analysis was performed to predict ground types up to ten segment rings ahead of the tunnel face using tunnel boring machine operational data. To achieve this, a hybrid model combining an autoregressive integrated moving average (ARIMA) model and a time delay neural network (TDNN) is proposed. First, the ARIMA model is used to predict machine data ten segment rings ahead of the tunnel face. Then, the predicted machine data is fed into the TDNN to predict ground types ahead of the tunnel face. We achieved a prediction accuracy of approximately 95%, which demonstrates the superiority of the proposed hybrid model. An engine was developed based on the proposed model to enable ground types prediction ten segment rings ahead of the tunnel face in any job site. Real-time geological risk management can be performed at a job site using the developed engine, thereby reducing and managing geological risks in advance.
Lastly, a method was studied to predict ground types ahead of the tunnel face utilizing operational data of the earth pressure-balanced (EPB) shield tunnel boring machine (TBM) when running through soil ground. The time series analysis model which was applicable to predict the mixed ground composed of soils and rocks was modified to be applicable to soil tunnels. Using the modified model, the feasibility on the choice of the soil conditioning materials dependent upon soil types was studied. To do this, a self-organizing map (SOM) clustering was performed. Firstly, it was confirmed that the ground types should be classified based on the percentage of 35% passing through the #200 sieve. Then, the possibility of predicting the ground types by employing the modified model, in which the TBM operational data were analyzed, was studied. The efficacy of the modified model is demonstrated by its 98% accuracy in predicting ground types ten rings ahead of the tunnel face. Especially, the average prediction accuracy was approximately 93% in areas where ground type variations occur.