The carboniferous igneous rock reservoir in the Dixi X well area contains a total of seven rock types, namely Basalt andesite, Basaltic volcanic breccia, Andesite volcanic breccia, Monzoporium, Tuff volcanic breccia, Granite porphyry, Rhyolite. The logging response characteristics are complex. The conventional intersection mapping method is difficult to distinguish between Basalt andesite and Basaltic volcanic breccia, Monzoporium and Tuff volcanic breccia, Granite porphyry and Rhyolite. ECS logging cannot effectively distinguish Granite porphyry and Rhyolite. Based on the technical principles and acquisition conditions of conventional logging, ECS logging and imaging logging data, and taking the lithology identification of thin sections as the benchmark, the lithology logging identification methods of igneous rock reservoirs were established by applying the intersection graph method, convolutional neural network method, "composition+acidity and alkalinity" method and "composition+structure" method. Firstly, a qualitative identification method for Andesite volcanic breccia, Basalt andesite and Basaltic volcanic breccia was established based on the intersection graph method and convolutional neural network method by conventional logging. At the same time, two lithological combinations were identified: the lithological combination of Monzoporium and Tuff volcanic breccia, and the lithological combination of Granite porphyry and Rhyolite. Secondly, the identification methods of Monzoporium and Tuff volcanic breccia were established based on the "composition+acidity and alkalinity" method by conventional logging and ECS logging. Finally, the identification methods of Granite porphyry and Rhyolite were established based on the "composition+structure" method by conventional logging and imaging logging. The results show that when the data of conventional logging, ECS logging and imaging logging are complete, the coincidence rate of identifying lithology by applying this method is about 86%. When there are conventional logging data and ECS logging data, Granite porphyry and Rhyolite can't be effectively identified, and the coincidence rate of lithology identification is about 66%. When there are conventional logging data and imaging logging data, it is impossible to effectively identify Monzoporium and Tuff volcanic breccia, and the coincidence rate of lithology identification is about 71%. When only conventional logging data are available, the coincidence rate of lithology identification is about 61%. It is recommended to measure ECS and imaging logging as much as possible to improve the coincidence rate of lithology identification.