基于MMAE-UKF的无人艇推进器故障诊断研究

打开文本图片集

关键词：水面无人艇；推进器；故障诊断；无迹卡尔曼滤波；多模型自适应估计；动态权重分配中图分类号：TN967-34；U672.7 文献标识码：A 文章编号：1004-373X（2026）06-0015-09

Research on USVpropeller faultdiagnosisbasedon MMAE-UKF

LIChanglong1，XIAOChangshi1，LIHaoxin1，LIUJiaxuan1，LIQiliang²，ZHOUChunhui1

1.Colf;i

Abstract：Afault diagnosis method basedonthecombinationof untraceable Kalman filter（UKF）and multi-modeladaptive estimation（MMAE）isproposedtoimprovethereliabilityofthepropulsionsystemofunmannedsurfacevehicles（USV）.Onthe basisofthetraditioalfilterdesign，teultiodeldaptieeiationisintrodued，andaUKFfultdiagnosisodelsforal leftthrusterfaultandrightthrusterfaultisestablished，respectively.Thelikelihoodofeachmodeliscalculatedbymeansof MMAE，andtheweightsareadjusteddynamicallyacording tothecurentstateoftheunmannedvehicletorealizethedetection andientificationoffaults.Thesimulationandreal-boattestingexperimentalresultsshowthattheproposedmetodcaffectiely identifyhefaulttypeafterthefaultoccurs，andundervariousmaneuveringconditionsoftheunmanedvehicle，itisstilableto detectthefaultsituationcontiuouslyandefectively，whichsignificantlyimprovestheaccuracyandsafetyofthepropulsionfault diagnosis. The results can provide a certain reference for fault monitoring of USV.

Keywords：surfaceunmanned vehicle;propeler;fault diagnosis;unscented Kalman filter;multi-model adaptiveestimation： dynamicweightallocation

0 引言

无人艇（UnmannedSurfaceVehicle，USV）是一种利用远程通信和自动导航等技术，在无人干预的情况下自主执行复杂海洋任务的智能平台。（剩余9832字）