Systems of seismic records processing and analysis using artificial intelligence

Summary<br>Heuristic approach towards the development of seismic data processing systems using artificial<br>intelligence implies the design of customized modules able to analyze raw seismic data and processing<br>results and use logic solutions to resolve seismic records processing, which are added to the existing data<br>conversion software. These logic modules accumulate modern knowledge on the physics of seismic wave<br>formation and propagation, as well as the knowledge on complex seismic data processing. New, higher class<br>of spatiotemporal convolution operators - non-linear logical operators, which could be adapted to the type of<br>seismic data processed - were introduced for mathematical formulation of “pure” processing procedures. At<br>the same time, all available a priori information could be included into the data processing procedure, which<br>becomes object-oriented. Special logical procedures could be engaged while on-line seismic records<br>processing. Though, this technique is most efficient while interactive expert assessment of seismic data. The<br>proposed data processing system is the most efficient in the areas of extremely difficult seismic and<br>geological conditions, when standard data processing software fails to produce high-quality results. Logical<br>procedures of expert seismic data processing are safe from false synchronism and undesirable dynamic<br>parameters distortion of seismic waves. Thus, the technique will be very helpful when geological objectives<br>are to be resolved by the interpretation of these parameters.