Abstract: Aiming at the problem of global navigation satellite system signal loss caused by building occlusion, multipath effect and insufficient satellite visibility, a precise point positioning (PPP) algorithm based on factor graph optimization is proposed for the integrated positioning of GNSS and INS. First, with reference to the classical PPP dual-frequency lonosphere-free model, the pseudo-range and carrier factors are constructed, and the nonlinear least-squares problem is solved according to the nonlinear optimization theory. Then, the optimized PPP location information is used as the GNSS PPP factor, and the refined pre-integration factor considering the rotation of the earth is constructed into the GNSS/INS loose combination factor graph frame, to realize the nonlinear optimization of integrated navigation information. The results of on-board real measurement show that the positioning accuracy of the proposed algorithm is 37.09%, 28.79% and 64.59% higher than that of the extended Kalman filter algorithm in the north, east, and down directions respectively for PPP; For GNSS/INS integrated navigation, the positioning accuracy of the algorithm is 49.08%, 41.22% and 71.86% higher than that of the extended Kalman filter algorithm in three directions.