Application of fuzzy pattern recognition method to gas logging data interpretation

The major purpose of compound logging is to prove the hydrocarbon potential of stratigraphic profile drilled and to reasonably interpret and evaluate the hydrocarbon-bearing layers found in accordance with display strength and componential composition data. In gas logging there are many factors influencing gas logging display, such as the physical property of hydrocarbons in formation, gas loss in mud outlet pipeline, mud property, the sensitivity of device to a certain component and the working effectiveness of degasser, etc. The influence of these factors on gas logging data is of randomness and fuzziness. On the basis of introducing the general methods commonly used for gas logging data interpretation, the basic principle of fuzzy pattern recognition method and the practical steps of how to interpret gas logging data by use of fuzzy pattern recognition method and relevant software developed are emphatically introduced in the paper. Through processing the practical data from a certain oil and gas field by this method, it is proved that its recognition rate is up to 81%. Therefore this method is relatively reliable for gas logging data interpretation.