Track Fusion And Prediction Of Target On A Moving Platform

Track fusion and prediction of target aims at fusing some dataaboutsome targetto obtaina more precisetargettrack, and establishing the appropriatemodel based on existed data to predict the positions of target in the follow-uptime. Track fusion and prediction of target on a moving platform is to achieve the goal of observing the target, fusing its track and predicting its next position in a mobile environment. It is a hot research direction to fuse and predict the track of target based on the background of moving platform in recent years. Whethermilitaryorcivilian areas, we needreliable and accuratetrackpredictions in the diversedevelopment environment.On amoving platform, because the platform and theobservation equipment installed on the platform are in motion, thesemeasurementdata obtained by the observation equipment, such asthe target azimuth and elevation, slant rangeand theplatform itselfposture, etc., can’t directly reflect thetargettrackininertialspaceandalso exist some errors more or less. In addition, the installation and calibration of the observation equipmentisa majorpractical engineeringproblem, because the calibration will directlyaffect theaccuracyofthe observation result. Furthermore, forthe track fusionandprediction of target,there are not publiclycomplete systemmodelcurrently.To solve these problems above, this paper, combined with practical project, byplatform decoupling, data pre-filtering, installation and calibrationcalculations of equipment,and then thetrack fusionandprediction of target, builds the corresponding mathematical model, in order toprepare for the future implementations of the real-time target observation, acquisition, trackingandpointingon amoving platform object. Itis quite valuable in engineering.Firstly, this paper introduces simply some of error sources involved in a moving platform,and describes the selected coordinate systems in detail. Then this paper introduces the coordinate relationship among the target, the observation equipmentand the platform,and achieves the transform between the inertialandnon-inertialcoordinate system, which is platform decoupling.Secondly, this paper gives a brief statement of the installation and calibrationof equipmentwhich must be solved in the practical projects. Thenchicken swarmoptimization(CSO)algorithm is applied, in this paper, to solvethis problem. This paper shows a detailed description of the basic structure of CSO algorithm, and makes improvements on the basic structure to calculate the approximatelyoptimalpostureandposition.Finally, this paper combines theinteracting multiple models(IMM)algorithm andleast squarecurve fittingalgorithm.Inthe local area, it uses the data sampled reversely at different intervals for many times tofit the close points and do the prediction, and takesmany of theprediction resultstothe fusionfor generating thetarget track. Globally, it adoptssliding window mode to finish extrapolationprediction step by step, and then achieves thetrackprediction of target.