Plot-Track Association And Attribute-Track Correlation In Data Fusion System

Plot-track association is a difficulty in data fusion system. To meet the need of large quantities of platforms in battlefield and the real-time performance of the fusion system, a new plot-track association algorithm-Approximate Multi-sensor Multi-target Joint Probabilistic Data Association (AMSJPDA) algorithm is presented in the paper. In the AMSJPDA, targets are divided into approximate clusters. Different from the cluster in MSJPDA, the approximate cluster has one "center target" and the other targets are its "neighbor targets" if there is at least one measurement inside the intersection of the validation gates of the center and the neighbor. Thus the number of targets in the same approximate cluster is not so large, which reduce the computational cost for the plot-track association. To compute the position estimation AMSJPDA takes into consideration all the measurements in the validation gate. Added all the measurements multiplied by their power, which is taken as the coefficient in computation, the near optimal estimation can be achieved. Compared with MSJPDA, AMSJPDA demands less time and has similar correct association rate. All suggest that AMSJPDA is suitable for the real-time fusion system in complex environment in clutter.Attribute-track correlation is also an important part in Data Fusion system. In simple simulation scene the traditional algorithm, which use azimuth as the basis of correlation, can work perfectly. But as the environment become more and more complex, the traditional correlation can't satisfy the demand of the system. To solve this problem, a new correlation algorithm-Multiple Factor Synthetic Decision Algorithm (MFSDA) is developed in the paper. In MFSDA, besides the azimuth the high accurate attribute parameters from ESM sensor is also considered as the precondition of attribute-track correlation. All the parameters are synthesized into a synthetic function, which is the core of the MFSDA. If the value of the synthetic function is under the validation gate, the correlation comes into existence. Use MFSDA the correct association rate is high and MFSDA also provides a new idea in colligation of azimuth and attribute parameters.