The invention discloses an estimation method for improving the measurement precision of an automatic ship identification system, and belongs to the technical field of sea ship state measurement. The method comprises the following steps: analyzing the model uncertainty and measurement noises based on a three-degree-freedom nominal kinematic model and a kinetic model of a ship on the water surface;establishing a three-degree-freedom disturbance model of the ship containing the model uncertainty and the measurement noises, and then establishing a three-degree-freedom measurement equation of theship on the water surface; designing a variable structure extended volume Kalman filter based on the measurement equation to estimate the motion state of the ship; tracking a target ship using an AISsystem; and sending the estimated ship motion state as a final target tracking state information through a communication unit. The filter of the present invention combines the advantages of a CKF filter and a sliding-mode variable structure filter, so that the accurate approximation of a two-dimensional system state information which contains the measurement noises under a Gaussian condition can be provided, and the filter has a good robustness characteristic on the uncertainty of the three-degree-freedom models of the ship on the water surface under a condition of random sea wave interference.