Deformation monitoring is a most direct and important approach to feedback engineering safety, mainly manual monitoring at present. But manual monitoring features slow information feedback and high operational risk. Problems of real-time data collection, monitoring information feedback and timely decision-making are more prominent especially during earthquakes or regional rainstorms. In view of this, we proposed a method integrating baseline edge difference and meteorological fusion correction and the TPS/GNSS fusion complementary monitoring scheme by researching the correction algorithms, monitoring schemes, field protection of instruments, automatic control and real-time acquisition for the rockfill dam of the gravel earth core of Pubugou Project. Moreover, we developed an integrated intelligent monitoring station system and a monitoring data acquisition and analysis system. The system has functions of self-adapting sampling frequency, self-optimizing observation period, self-identifying emergency scene, self-matching optimal model, and self-protecting instrument. The system significantly improves the quality of deformation monitoring data, enhances the intelligence level of deformation monitoring, and provides a feasible reference solution for automatic deformation monitoring.