Study Of Data Processing Algorithm Based On The Surface Movement Radar

In recent20years,the global transportation of aviation developed very fast,the size of theairline is increasingly growing,to improve the ability of the airport surface movement,SurfaceMovement Radar (SMR) is an effective tool for rational programming ground traffic control onairport area,and is widely applied in operation of the oversea and domestic large airports. Dataprocessing module of a certain SMR simulation platform has been studied,which plays an importantrole in the whole platform.In this paper the relevant algorithms about associating,filtering andmaneuvering target tracking have been designed and simulated.Firstly,we introduce target tracking problem in the interferential enviroment which needs todiscuss data association algorithm.Classical data association algorithms are introduced and theiradvantages and disadvantages are summerized.JPDA algorithm is specially analyzed and simulatedin detail.Secondly,the common classical filtering methods in the target tracking system areintroduced.Due to the lack of online adjustment ability,UKF algorithm can’t track saltation states.Inorder to overcome the limitations of UKF,an improved MA-STUKF algorithm is developed, whichis based on memory attenuating filter and stong tracking filter(STF)，to adjust a filtering gain matrixonline by introducing a time-varied fading matrix.The simulation results demonstrate that theimproved algorithm has better maneuvering targets tracking performances and target statesestimation precision.Finally,due to the complexity and uncertainty of the target movement,it is hard to representedaccurately by single arithmetic expression.Interacting multiple model (IMM) algorithm has a bettertracking effect than single model filter algorithm because their inputs and outputs of each filter areweighted comprehensively.An improved IMM algorithm is developed combining UKF algorithmwith markov parameter adaptive technology to solve the problem of non-linear filter formaneuvering target.The transition probability of the classical IMM estimators is determined byexperience,but in proposed algorithm the markov transition probability is adaptively adjusted to areasonable distribution.