Classifying rock masses using artificial neural networks

The primary purpose of the Rock Mass Classification methodologies is to classify rock masses into behaviorally and mechanically homogeneous categories. A variety of classification methods such as Barton's (Q) and Bieniawski's Rock Mass Rating (RMR) have been used. This paper uses a feed-forward backpropagation-type Artificial Neural Network (ANN), applied to data from tunnels in Greece. Results show that with a smaller number of input variables the ANN can place a rock mass in the aforementioned classification ratings very quickly and with very high accuracy. Further, the engineering bias often present in the traditional ways of achieving these ratings is reduced.