Cross-borehole geological interpretation model based on a fuzzy neural network and geotomography

摘要

This paper demonstrates a method of predicting sub- surface geological characteristics in the region between boreholes using a fuzzy neural network(FNN), inverted geophysical data(geotomogram), and fuzzy geologi- cal knowledge. Traditionally, geological interpretations are often of inadequate reliability in cross-borehole re- gions because of the scarcity of quantitative data. Hence, geological interpretation has been a subjective process. Although there is a paucity of data, a fair amount of in- formation exists in the form of human perception, sum- marized as either "rough maps" or "vague descriptions".