| With the popularization and application of GPS,AIS and sensors,the acquisition of trajectory data is becoming convenient increasingly.In the process of trajectory data acquisition,it is difficult to avoid contact abnormality of equipment,communication malfunction,congestion in information transmission network and other phenomena,so the problem of missing trajectory is still serious.To a certain extent,the missing trajectory will affect the analysis of the trajectory and make wrong conclusions or judgments easily.It is very important to estimate the missing trajectory correctly.But the missing trajectory is random,the motion state in the missing trajectory section is difficult to estimate,which increases the difficulty of trajectory restoration.This thesis found the difficulties and problems to be solved in the research of trajectory restoration algorithm,which through summarizing and analyzing the existing trajectory restoration algorithms.Regarding the above problems,missing trajectory recognition algorithm and trajectory restoration algorithm based on vector interpolation are proposed.The specific research content are as followings:1.A missing trajectory recognition algorithm is proposed.Firstly,the distance between two adjacent points in the trajectory is calculated,and compares whether the distance is greater than or equal to the preset distance threshold.Secondly,the time difference between the two adjacent points is calculated,and compares whether the distance is greater than or equal to the preset time threshold.Finally,if the condition of greater than or equal to is satisfied at the same time,it can be determined that the trajectory contains missing,output the recognition result,as well as the dynamic information of two adjacent points.2.An algorithm for trajectory restoration based on vector interpolation is proposed.Firstly,the concept of bidirectional iteration is proposed.The algorithm interpolates the dynamic information by bidirectional iteration and two predictive trajectories are obtained.In the whole iteration process,the dynamic information of each point is fully utilized,which can avoid excessive dependence on the dynamic information of the starting and ending points of missing trajectory segments,resulting in excessive deviation of the intermediate point.Secondly,weighted average of the two predicted trajectories is used to obtain the final repair trajectory.Finally,through the experiments of restoring missing trajectories in different motion states,and comparing with other trajectory restoration algorithms,the effectiveness of the proposed algorithm is verified. |
