The design and implementation of dip arrow plot pattern recognition system

Dip logging is a method of well logging. From the raw data obtained by dip logging, a dip arrow plot can be obtained using the relation comparison method of pattern recognition. The data tend to be very much scattered and random. The design principle and method of an expert dip arrow plot pattern recognition system are described. It incorporates a mathematical model method and an artificial intelligence technique. An angle processing algorithm and an ordered sample grouping algorithm are proposed. This system has been implemented on an IBM-PC/XT and on compatible machines. Its classification rate is 98%, obtained by processing 21 data patterns from the Zhongyuan oil field in China.<<ETX>>