Modeling tunnel boring machine performance by neuro-fuzzy methods

This paper presents the results of a study into the application of neuro-fu zzy methods to model the performance of tunnel boring machines. A database consisting of over 640 TBM projects in rock has been used. It is shown that neuro-fuzzy methods give better results than other, more conventional, mod eling approaches. Fuzzy set theory, fuzzy logic and neural networks techniq ues seem very well suited for typical geological engineering applications. In conjunction with statistics and conventional mathematical methods, hybri d models can be developed that may prove a step forward in the practice of ground engineering. (C) 2000 Published by Elsevier Science Ltd. All rights reserved.