A Study In Track Loss By The Method Of Twice Track Association And Fusion

In actual sea battle-filed, because of geographical environment, EMI, electronic silent, etc.,problems as sensor signal interrupt, target loss, track discontinuity appears. The fusion centercannot always get the data of situation elements continuously. In this case, we need a trackassociation algorithm for track loss coursed by long time interrupt.To start from the data fusion, this paper introduces the concept, model and development ofdata fusion. Then it describes the issue of track loss, and puts forward a track association andfusion algorithm for track loss by studying the concept and study of track association and trackfusion.The algorithm is composed by two association processes: first a track association algorithmand then a track association and fusion algorithm. The first track association is an improvedfuzzy track association algorithm. In the situation of signal interrupt and target loss, thisalgorithm uses the method of forward/backward forecast to avoid incorrect association. Thenthis algorithm can associate the breakpoints of the interruption and display a continuous trackby using the methods of uniting speed fuzzy factor and course fuzzy factor, improvingacceleration fuzzy factor, twice association strategy and autoregulation association thresholdvalue of membership function. Then the second track association and fusion can accomplish thetrack association and fusion by a track association algorithm based on local shape features.Considering the unstable changes of tracks, this algorithm changes this situation into a vectorsof figure characteristic, in order to comparing the similarity of two tracks. And then itaccomplishes the track association and fusion.At last, this paper puts forward a simulation. The result shows that in the situation of signalinterrupt and complex track, this algorithm has a good effect. However, in case of more than5record periods interrupt time and close and cross tracks, this algorithm is not good enough.These problems will be solved in further research.