Abstract:The technique for reservoir pattern recognition involves the following procedure:·Data processing,including preprocessing,noise elimination,various corrections,and the removal of influence factors which are not related to objective interval.·Characteristic pickup ,followed by the computation of nhydrocarbon characteristic indicators,which are used to form n-dimensional characteristic vector.·Computing,from classificator,the discriminant function value for classifying the n-dimensional characteristic vector.The function value higher than zero belongs in classification I,relating to hydrocarbon reservoir;and the value less than zero belongs in classification II,correlating to the formations which bear no hydrocarbon.We also detail how to design the classificator and how to pick up the characteristics.Forward modeling data and two examples are given in the paper.