Restoring AIS trajectory of inland waterway ships method based on navigation experience

There exists severe AIS signal loss in inland navigation environment, and the ship trajectory has characteristics such as long sampling interval, sparsity, and discontinuity. Traditional interpolation methods are mainly applicable to the trajectory restoring of the straight channel, which cannot be applied to the inland channel with curved and complicated topography, for the curve motion characteristics between adjacent trajectory points are easy to be ignored. In this paper, a method for the restoring of inland ship trajectories based on the navigation experience is proposed. Firstly, the historical trajectory records of the other ships in the neighborhood of the sampling point are extracted. Secondly, the navigation experience of the other ships is modeled based on the multi-parameter constraints of time, location, course etc., and kernel density estimation is used to acquire the heat map of ship navigation. Lastly, the highest probability is searched based on the ship heat map to figure out the restored trajectory. Two typical inland channels are also presented as the experimental scenarios, and a comparison is made between the restoring results using the proposed method and the spline function method respectively. The comparison results demonstrate this method can better apply to the curve and complicated topography of inland channel, and can quickly restore and reconstruct the sparse inland trajectory.