The Data Fusion Of Radar And AIS And The Performance Evaluation Of Radar

With the development of navigation technology, more and more Marinecommunication navigation equipment are produced to provide strong support forsafe navigation and protect the marine environment. Traditionally, radar, ARPA(Automatic Radar Plotting Aid) and VHF(Very high frequency) radio were used toget information during sailing. AIS(Automatic Identification System) wasdeveloped based on VHF DSC in the late1990s, which used GPS to providelocation information. AIS can send out inquiry signals to the surrounding ships andexchange information automatically with VTS(Vessel Traffic Service) which is usedto get the information and the state of vessels. AIS can remedy the radar’s defaultof range finding, so, it is widely used. Because of the different precision, whenradar and AIS are used in the same area of ocean, their data are associated andredundant, so, in order to increase data efficiency, data fusion based on these twosensors is very necessary, which can be used to obtain the information more reliableand complete.In this paper, there are two very important key problems, one is data fusion andthe other is the evaluation of radar performance. The research object of this thesis isHF GWR (High Frequency ground wave radar). This paper put emphasis on theposition level of data fusion. The evaluation of radar performance is to assess theindicators of radar’s distance, bearing, heading, velocity, and radial velocity. Fisrtly,this paper briefly introduced the performance features of the radar and AIS.Secondly, the paper chose a data fusion algorithm and realized it by MATLAB.Thirdly, the paper designed a software by VC++, which can fuse data and evaluatethe performance of the radar, and also provided visualization capabilities.The most important thing of data fusion is track correlation. The paper used adata-fusion algorithm based on vague comprehensive approach. Radar locates atarget with its distance and bearing when AIS locates it with its latitude andlongitude. So, coordinate transformation should be used to make their locationinformation into the same coordinate system. At the same time, the date rate isdifferent between radar and AIS, then the time of location information of the sametarget is also different. So the information from the two sensors should be synchronized in order to judge whether the tracks from the two sensors are related.Distance, heading and velocity are the main factors which can impact the measuresof the ship’s location, so the paper put forward a compositive approach of similaritycomputation by these three factors for track correlation. At last, the paper usedweight fusion algorithm to fuse the related track data of AIS and radar, and alsoused the related track data to evaluate the performance of the HF GWR.