Dynamic modeling for route-based motions

摘要

This paper is concerned with the modeling problem of route-based motions. A typical route-based motion - straight-line motion - is modeled. Distinct from the conventional bottom-up strategy of the general motion modeling, the proposed modeling adopts a combination of top-bottom and bottom-up strategies. The process noises of the route-based motion model is discovered to be negatively correlated rather than the conventionally assumed white, which is a significant characteristic of such motions. Performance of the route-aided tracking algorithms is evaluated on an air traffic control tracking scenario. Simulation results demonstrate their superiority over the classical target tracking algorithm in terms of accuracy.