Application of Self-Organizing Competitive Network in Lithologic Identification of the Logging Data.

The geological information of logging data is very important for people to determine oil reserves and make the plan of exploitation. So it is essential to identify litho logy of the logging data. Neural network with self-organizing, self-learning and the ability of highly non-linear mapping has been widely used in the field of classification. It has achieved good results. Using self-organizing and self-learning ability of self-organizing neural network, this paper analyzes the factor of litho logic identification, establishes self-organizing competitive network model based on MATLAB. By comparing the two structures of basic competitive network and self-organizing competitive network we achieve litho logy classification. Experimental results show that it is feasible to identify litho logy of the logging data by self-organizing network model. It is a new method of litho logic identification and Its correct rate is high. [ABSTRACT FROM PUBLISHER]