Research On Multi-target Trajectory Fusion Technology Based On Sensor

Trajectory fusion is an important part of target tracking,which fuses the same target trajectory together by fusion technology to make the measured trajectory closer to the real trajectory.With the rapid development of modern science and technology,the requirement of trajectory fusion accuracy is getting higher and higher.In addition,when tracking multiple moving targets in complex environments,the quality of trajectory correlation algorithm will directly affect the effect of trajectory fusion,and the quality of trajectory fusion algorithm will directly affect the accuracy and efficiency of target tracking.In view of this,based on the analysis of existing target tracking algorithms,this thesis focuses on the multi-target trajectory correlation and trajectory fusion technology.In view of the shortcomings of existing algorithms,adaptive trajectory correlation algorithm and weighted trajectory fusion algorithm based on local state estimation are proposed to improve the prediction accuracy of trajectory fusion.The achievements in this thesis mainly include the following aspects:Firstly,an adaptive trajectory correlation algorithm is proposed to improve the accuracy and reduce the computational complexity of trajectory correlation.Considering the uncertainty of multi-objective trajectory state information,the algorithm chooses the corresponding trajectory correlation algorithm through multi-objective trajectory correlation strategy to solve the problem of large amount of computation caused by the mismatch between trajectory state and the selection of correlation algorithm.At the same time,aiming at the inaccuracy of the existing double-threshold trajectory Association algorithm,a speed-based double-threshold trajectory association algorithm is proposed,which improves the accuracy of trajectory Association in multi-target tracking.Secondly,a weighted trajectory fusion algorithm based on local state estimation is proposed to deal with the error caused by the change of sensor accuracy in the data processing of existing trajectory fusion algorithms.the algorithm uses the local state estimation of each sensor to complete the weighted combined filtering of multiple sensors,and the error problem caused by the change of sensor accuracy is solved,thereby improving the accuracy of traj ectory fusion.Finally,the multi-target trajectory fusion system based on radar sensor is simulated by using the radar transmitting and receiving signal mechanism in SystemVue platform,and the above results are simulated and tested.The simulation results show that,compared with the existing algorithms,the proposed algorithm improves the accuracy of trajectory correlation and fusion in the multi-target trajectory fusion system model,and achieves the purpose of improving the tracking accuracy of the target.