TBM（Full Face Rock Tunnel Boring Machine）collects a large amount of data in the excavation process. However, due to the mechanism of PLC data collection, the original data is large-scale and contains miscellaneous information. Based on this, through the standardized data preprocessing, this paper preliminarily divides the original data into different boring section by the threshold setting and standard deviation division, and then carries out noise reduction and filtering, classifies the abnormal data to improve the quality of data. Based on the processed data, this paper analyzes the characteristic parameters field penetration index（FPI）and torque penetration index（TPI）, studies the correlation between characteristic parameters and surrounding rock geological conditions. The results show that the data preprocessing program can better divide the boring section. The characteristic parameters have a high correlation with the geological conditions of surrounding rock. It can preliminarily judge the abnormal working conditions encountered by TBM such as stuck machine and weak surrounding rock through the threshold.