METHOD AND DEVICE FOR PREDICTING THE SPATIAL DISTRIBUTION OF GEOLOGICAL STRUCTURE TO PREPARE GEOLOGICAL MAP

摘要

PROBLEM TO BE SOLVED: To precisely predict the spatial distribution of a geological structure with less data and prepare an geological map by subjecting geotomogram and fuzzy soft map data by hierarchical neural network(NN) processing. SOLUTION: In the geological inspection in a field, data are inputted from the boring data input part 13 of an input part 10 to provide a geological map. With respect to the data extracted from the bore hole, the fuzzy map forming part 17 of a fuzzy NN part 11 forms a soft map data every geological section G1-G8. A hierarchical NN part to which this data is inputted also receives the geotomogram (seismic wave speed, specific resistance) obtained by physical probe from a geotomogram input part 14, and forms a fuzzy NN model on the basis of all the data. The geological map and the prediction reliability are outputted from a geological map output part 20 and quantitatively from a prediction reliability output part 21 through a non-fuzzy part 19, and fuzzy geological features every type are directly outputted to a fuzzy geological map output part 22.