基于UKF的水下装备轨迹估计滤波算法

打开文本图片集

中图分类号：P715.9 文献标志码：A 文章编号：105-9857（2025）01-052-09

（1.ScholofOceanandEarthScience，TongjiUniversity，Shanghai，China; 2.ScholofElectronicandInformationEnginering，TongjiUniversity，Shanghai，China）

Abstract：TheunderwaterequipmentcariedbytheEastChinaSeaseabedobservationnetwork isoftenlostduetohumanactivities.Whensolvingthetrajectoryofunderwaterequipmentbyinertialnavigation，thenoisecausedbythecomplexoceanenvironmentandsudenmotionleads tolargecalculationerors.Aimingatthelimitationthattheclasicunscented Kalmanfilter nedsacuratenoisemodelanddynamicmodeltooutputbeterfilteringresultsinthetrajectory estimationproces，anadaptivefilteringalgorithmbasedonSage-HusaunscentedKalmanfilter isproposedforsolvingthetrajectoryofunderwaterequipmentbeingtowed.Firstly，theimplementationproces oftheclasicunscentedKalmanfilterisexplained，andtheideaofSage-Husa adaptiveadjustmentofnoiseisintroduced;then，onthisbasis，thepredictionresidualvectoris introducedtoreducetheinfluenceofgros eroronthefilteringresult，andatrackingfactoris introducedtoimrovetheadatabilit ofthefiltertosudenmotion;final ，theefectivenesof thealgorithmisverifiedbysimulationexperimentsandtowingexperiments.Theexperimentalresults showthatheadaptivefilteringalgorithmbasedonSage-HusaunscentedKalmanfilterefectively reducesthedivergenceerorcausedbytheunderwateracousticenvironmentandsudenmotion， andimprovesthepositioningacuracyandpositioningstabilityofunderwaterequipmentwhen beingtowed.

Keywords：Unscentedkalmanfilter，Adaptivefilteringalgorithm，Underwaterequipmenttowing，Advancedalgorithm

0 引言

随着科技的发展和建设海洋强国战略目标的提出，东海海底观测网作为我国首个智能化、一体化、开放式的海底观测网，现已取得良好的成果[1]。（剩余11068字）