Multi-sensor Vessel Target Tracking And Fusion Algorithm

Currently,the Automatic Identification System(AIS)and High Frequency Surface Wave Radar(HFSWR)is two major sensor equipment for a wide range of continuous monitoring of marine vessels.With the help of AIS,accurate information about the cooperating vessel can be obtained.HFSWR can continuously monitor all vessels in a large area under all weather conditions.More accurate and abundant information about vessel activities can be obtained from the fusion data of AIS and HFSWR.Aiming at the problem of tracking and fusion when AIS and HFSWR work together,this thesis focuses on HFSWR track generation and repair,AIS and HFSWR dual-frequency track fusion.Finally,a marine multi-sensor data fusion module for engineering applications is designed.The main research content is divided into the following points:1.According to the sensor measurement characteristics of HFSWR and the motion characteristics of ships at sea,a track generation and repair algorithm for HFSWR is studied.Based on the detection target points accumulated by the radar in multiple frames,the M/N logic method is used to the track initialization.The minimum cost algorithm is used for the track-to-point association of the successfully started tracks.Finally,the Debiased Converted Measurement Kalman Filter(DCMKF)is used for state filtering.For part of the track breakage generated in the track generation,a reverse test is used to determine whether it is from the same target.The experimental results show that the proposed track generation and repair algorithm has reliable track generation and repair capabilities.2.In response to the breakage of the radar detection tracks,an algorithm is designed to track fusion for AIS and HFSWR dual-frequency tracks.The propensity sequence between tracks is obtained by calculating the fuzzy comprehensive evaluation value between tracks,and then the tracks are clustered by an iterative search algorithm to obtain track clusters.Finally,the track clusters and propensity sequences are input to the Improved Gale-Shapley(IGS)algorithm.IGS algorithm is used to carry out the game to obtain the tracks association result.The associated track is fused in the public time,and the track deviation is corrected in the non-public time.The measured data and simulated data of dual-frequency HFSWR and AIS are used for experimental tests.The experimental results show that the designed algorithm can deal with the track break problem in track-to-track association ignored by other algorithms,and the association performance in dense areas is better than the global nearest neighbor algorithm.3.According to the actual needs of engineering applications,the shipborne multi-source fusion module is designed based on the C# programming language.The main functions are target points fusion and tracks fusion.The module has the functions such as real-time reception of local sensor data,screening of data files with fusion conditions,and data fusion processing.After verification,the module can meet the application needs in the actual engineering environment.