The western side of the Sichuan Basin is one of the regions with the highest frequency of extreme heavy precipitation events (HPEs) in China. However, due to the complexity of terrain and atmospheric circulations systems, the formation mechanism of HPEs remains elusive. Based on the gauged data by China Meteorological Administration from 2001 to 2020, GPM-IMERG precipitation and ERA5 reanalysis data, we selected 100 extreme HPEs in the west of Sichuan Basin, and classified these HPEs into three categories with K-means clustering method. Then, the anomalies of atmospheric circulation and its evolution with respect to different category of HPEs are explored. The results show that, during the period of precipitation, the geopotential height is characterized by a positive anomaly at the upper-level (200hPa) and negative anomaly at the lower-level (850hPa), together with the enhanced vertical wind speed, a "top-cold and bottom-warm" atmospheric temperature structure, and the increased water vapor transport departing from the low latitude ocean area. However, their atmospheric circulations configuration, including the South Asian High, the Western Pacific Subtropical High, and the westerly jet in upper level, shows obvious difference and plays a dominant role in in shaping the precipitation formation of different types and associated water vapor transport: For category 1 HPEs ranked the highest frequency of occurrence, its associated water vapor transport mainly comes from the Bay of Bengal and the South China Sea, which controlled both by the Indian monsoon and the East Asian monsoon. As to category 1, HPEs occurs with the strengthened WPSH, and its water vapor mainly comes from the South China Sea, while the water vapor transport from the Bay of Bengal is restricted. When it comes to the category 3, its water vapor transport mainly controlled by the East Asian monsoon. Prior to the occurrence of HPEs, it can be found that the Rossby wave action shows an increasing trend, the atmospheric vertical wind speed increase significantly, and the temperature anomaly was more significant. These features are conducive to the occurrence of extreme precipitation events. The results also indicates that the changes in WPSH position and intensity on synoptic time scales are better indicators for the prediction of heavy precipitation formation compared to SAH.