To reduce the impact of platform magnetic interference and sensor errors（triaxial non-orthogonal errors，triaxial sensitivity inconsistency errors and zero bias errors）on the triaxial fluxgate sensor mounted on the underwater platform，an integrated calibration method based on improved particle swarm optimization algorithm is proposed in this paper. On the basis of analyzing the error source，a mathematical model for error calibration is established，and the validity of the calibration method is verified by two simulation examples. The simulation results show that compared with the traditional particle swarm optimization algorithm，the improved algorithm has higher robustness against noise and solution accuracy. After calibration，the measurement errors caused by the sensor errors and platform magnetic interference are effectively suppressed.