A Neural Network Approach to Comprehensive Classification of Rock Stability, Blastability and Drillability

Xiating Feng
DOI: https://doi.org/10.1080/09208119508964719
1995-01-01
Abstract:ABSTRACT Comprehensive classification of rock masses, an important activity during exploration, design and construction for underground openings, is restrained by our limitations in defining the complex geological environments and in modeling the relationships between geological features and rockmass behaviors. In this paper we present a novel approach for this problem. Our approach makes use of the ability of learning, generalizing and nonlinear dynamic processing of neural network. The knowledge of rockmass classification is first learned from case histories using two multilayer feedforward neural networks and an improved Back-Propagation algorithm and then stored for the use of the neural network in recognizing new cases. The adaptive recognition of new cases shows this approach is effective.
What problem does this paper attempt to address?