Stratigraphic Method Based On Artificial Neural Network Faults On Both Sides Of Median Follow-up Study

The layer tracking between the fault's two sides plays an important role in the seismic data interpretation. As the major trait of geological formation, the fault acts the aggregation function during the course of transportation and storage of the petroleum and gas. The traditional method of layer tracking and location between the fault's two sides uses the characteristic similarity of wave form to track and locate it artificially. This method relies on the artificial experience, and lack of the quantitative analysis method. Concerning this issue, and based on the characteristic similarity of reflective layer, this thesis creatively presents an automatic method of layer locating and consistency tracking by the Artificial Neural Network.This thesis specifies the basic theory and network structure of Artificial Neural Network; concerning the layer tracking between the fault's two sides, the thesis extracts the characteristic parameter of the real seismic data by the traditional method of parameter extracting. In terms of the event picking and tracking, it compares the obliquity sector method and edge detection method, then provides the SOFM method which is good at classification to locate the event. Finally, this thesis uses the BP network track the layer of between fault's two sides.According to the processing and analysis for the real seismic data, the method based on the SOFM and BP algorithm in the Artificial Neural Network is efficacious in the layer tracking and location between the fault's two sides. It is more precise than the traditional method, and presents a automatic and quantitative method for the layer tracking and location between the fault's two sides.