Research On Data Fusion Based On Netted Radar To Improve Measurement Precision Of Aerial Targets’ Flight Track

Data fusion of netted radar system is one of the typical applications of multi-sensordata fusion techniques. Compared to measuring targets with a single radar, properlyfusing the detection data from multiple radars can acquire more precise and completeinformation about targets.Systematically summarized at first in the thesis is the research status quo home andabroad of multi-sensor data fusion and radar networking technology. Then multi-radardata fusion precision is quantitatively analyzed under both circumstances of usingequal-precision radars and non-equal-precison radars after carrying out the study onhow to improve the data fusion precision of netted radar system. Established later ismulti-radar deployment model, which and its influence on the precision of data fusionare both studied. And the corresponding relationship between the assessment indices ofmulti-radar deployment and data fusion precision is also formulated.The thesis focuses on the study of how to improve the measurement precision ofaerial target tracks through real-time data fusion of netted radar system, discusses twodata fusion algorithms, Between-Data Support Degree Matrix Estimation (BDSDME)and LMS Adaptive Transversal Filter (LATF), and advances a new data fusionalgorithm, Kalman Filtering Measurement Value Substitution (KFMVS). Simulationsshow that all the three algorithms can respectively improve the measurement precisionof target tracks in varying degree. Nevertheless, KFMVS, advanced by the author, takesadvantage over the other two in real-time data processing of netted radar system. Ittakes the output of BDSDME as Kalman Filter’s input, and, without benchmark dataand a priori knowledge about whether measurement value being subject to normaldistribution or not, is able to improve the data fusion precision by adjusting the weightsin real time.Then, combined with the application requirements of the equipments in TestCenter, the thesis utilizes a preprocessing technique suitable for multi-radarmeasurement data, which includes spatial and temporal registration, deviation errordisposal and data fusion, implements unilateral and bilateral deployment respectively inboth two-and three-different-priciosned radar scenarios, and quantitatively analyzes theeffect of radar networking upon improving the measurement precision of aerial targettrack by adopting each one of the three above-mentioned algorithms.Research results show that Test Center’s radar equipments can acquire morecomplete and precise information on targets through reasonable deployment of radarsand data fusion. The work done in the thesis not only provides valuable reference fornetworking of radars and for enhancing data fusion precision, but also are of greatimportance to improve both data processing and assessment level and equipments’measurement capability in Test Center.