Vortriebssynchrone Prognose der Setzungen bei Flüssigkeitsschildvortrieben auf Basis der Auswertung der Betriebsdaten mit Hilfe eines Neuro-Fuzzy-Systems (English shop)

Preview

Table of Contents, Datei (15 KB)
Extract, Datei (270 KB)

This dissertation investigates the prediction of ground and surface settlements in mechanized tunnelling.
The research focuses on tunnel excavation using slurry shield tunnel boring machines.
Its objective is to predict tunnelling-induced settlements during ongoing excavation.
A novel prediction method based on a neuro-fuzzy system is developed for this purpose.
The approach combines fuzzy logic with artificial neural networks.
Operational data recorded during tunnel excavation provide the basis for the analysis.
The interactions between shield tunnelling and ground movements are investigated.
Particular attention is given to tunnel face support and annular gap grouting.
Excess soil excavation and geotechnical influencing factors are also considered.
Fuzzy rule-based systems enable the representation of complex and uncertain relationships.
Artificial neural networks are used to adaptively improve the system’s knowledge bases.
The model considers the temporal and spatial development of the settlement trough.
The proposed method is evaluated using data from two real tunnel construction projects.
The results illustrate the potential and limitations of data-driven settlement prediction.
The dissertation contributes to improved process monitoring and settlement minimization in tunnelling.