Abstract: The principle of PNN (probabilistic neural network) method and its algorithm training processes, especially the selection of lithologic parameters and establishment of lithologic identification model, are discussed in detail in the paper. The advantages and disadvantages of different identification methods are analyzed with comparison between PNN and other six lithology identifications, as well as the prediction results under the same condition. The study shows that the PNN method is the best in production and application with the shortest training identification time. The use of artificial intelligence neural network for automatic interpretation and analysis of logging data can meet the needs of timeliness and rapid interpretation of logging while drilling (LWD).