Research On Data Processing Technology Of Ship Supervision Radar

The supervision of ship targets on the surface has always been a key research direction,which is of great significance to navigation safety.Navigation radar and AIS(Automatic Identification System)are two commonly used navigation AIDS,which can ensure the navigation safety of ships to a certain extent.Among them,navigation radar is an active detection device,which can still give the movement information of the target to guide the ship to avoid collision and maneuver when the target does not cooperate.AIS needs to passively receive the broadcast of the target to obtain the target position,and the detection modes and types of information obtained by the two devices are highly complementary.Combined with the characteristics of the two,this paper studies the core technology of ship supervision radar data processing.The specific research contents are as follows:1.Firstly,this paper studies the data processing technology of sea target tracking of navigation radar.A DBSCAN(Density-Based Spatial Clustering of Applications with Noise)condensation algorithm assisted by Doppler velocity information is proposed to calculate the shape information of the extended target.Based on the fact that the Doppler velocity of the same extended target is roughly the same,the algorithm not only uses DBSCAN to judge the range neighborhood,but also evaluates the doppler velocity threshold,limiting the condition that the point trace becomes the core point,and enhancing the ability to distinguish the adjacent extended target.For target tracking problem under the sea clutter interference,a kind of sea surface target tracking algorithm is studied in this framework,using the association class speed advantage of the algorithm,using GM-PHD(Gaussian Mixture Probability Hypothesis Density)algorithm to filter out noise,reduce the clutter density before using JPDA(Joint Probabilistic Data Association)algorithm on data association,at the same time,the application of the characteristic information of the target tracking algorithm framework,And eventually form an accurate track.The above three methods are connected in series to complete the sea target tracking of navigation radar.Aiming at the problem of track management of sea target,the track management logic is modified to reduce the generation of false track and make it more suitable for multi-target tracking on sea surface.The feasibility of the radar target tracking algorithm is verified by simulation experiments.2.Secondly,this paper studies the data fusion of radar and AIS,and analyzes the space-time alignment of radar data and AIS data in the data fusion.In order to reduce unnecessary track correlation calculation,T-D(Time-Distance)rough correlation criterion is used to carry out rough correlation between the two tracks,which improves the computational efficiency.Considering the fuzziness between tracks,a multi-factor fuzzy track correlation algorithm based on fuzzy mathematics is studied,which uses membership degree to describe the similarity between tracks.In terms of track fusion,based on the minimum mean square error of the system,weights are calculated for the associated track pairs respectively,and the statistical weighting method is used to complete the track fusion,which not only reduces the amount of calculation but also completes the track fusion of radar and AIS.Finally,the effectiveness of the algorithm is verified by simulation.3.Finally,this paper uses the algorithm and its improvement mentioned above to design the data processing software of ship supervision radar combined with the requirements of sea surface target supervision.The software includes radar data processing module,radar and AIS track fusion module and output display module,with real-time radar data processing,radar and AIS data fusion,track display and other functions.