Abstract: 【Objective】The Wufeng-Longmaxi formations in the Yuxi Block of the Sichuan Basin are typical shale gas reservoirs in China. The strong interbedded heterogeneity of these formations leads to a significant amount of redundancy in the collected logging curve information and complex coupling relationships between curves, resulting in the difficulty and low accuracy of lithofacies identification using traditional methods, which urgently needs technical innovation. 【Methods】In this study, based on lithofacies classification and analysis, an intelligent lithofacies identification method combining Principal Component Analysis (PCA) and Random Forest algorithm was developed. 【Results】The research findings are as follows: Firstly, by optimizing the logging curves using PCA, the deep coupling of logging curves can be achieved, thereby reducing the impact of factors such as redundancy of logging information and complex coupling relationships between curves on lithofacies identification, resulting in more scientifically effective data information. Secondly, by adding minor variations to the original data without altering its lithofacies, data augmentation can be achieved, addressing to some extent the problem of poor generalization ability and stability of the model caused by small or imbalanced datasets when using the Random Forest algorithm. Finally, the intelligent lithofacies identification method constructed by combining PCA and Random Forest algorithm achieved an identification accuracy of over 83%, demonstrating its high precision and strong applicability. 【Conclusion】This method not only overcomes the difficulties in lithofacies identification in the study area to a certain extent but also greatly improves the efficiency of lithofacies identification, which is of great significance for promoting the economic and efficient development of shale gas in the study area.