简介:海军和另外的国防部组织逐渐地为许多使命和应用对无人的表面车辆(USV)的使用感兴趣。术语USV指没有一个全体乘务员,操作水的表面的任何车辆。USV有潜力,并且在一些情况中表明的能力到到有人驾驶的力量的还原剂风险,提供必要力量增加完成军事使命,执行有人驾驶的车辆的任务不能,并且那么以一个方法,那为海军是买得起的。世界范围的USV活动以及USV的一般技术挑战的调查下面被介绍。对USV的一般描述与他们的典型应用一起被提供。开发USV的技术挑战包括它的智力水平,控制,高稳定性,和发展费用减小。通过研究人员的联合努力在全世界,USV的发展在不久的将来将进入一个新阶段,,,这被相信USV能很快在军事、平民的服务两个都广泛地被使用。
简介:Inthispaper,maximum-likelihood(ML)anditsrelaxationalgorithm,whichareusedtoidentifythemathematicsmodelofanunderwatervehicle(UV),arcdiscussed.Withthetrialdataofzigzagtests,thehydrodynamicderivativesoftheUVwereestimated,andtherelaxationalgorithmisconfirmedtohavebetterastringencyfromthecontrastbetweenthetwomethods.Thenasimulationenvironmentbasedontheseparametersisestablishedtoverifythevalidityandeffectofthesemeth-ods.Theresultshowsthemodeliscredibleandthemethodsareveryusefulfortheresearchofmaneuverabilityandadaptivecontrolofunderwatervehicles.
简介:Itisanimportantcontrolprocesstooperatemotionofansubmergencerescuevehicle(SRV).Seeingthatthemotionofthesubmergencerescuevehicleisspecial,itisnecessarytoemploynon-linearpredictivecontrolsystem.Forthisreason,continuousdynamicperformanceofthesystem,thelogicalcomponentsandtheoperativerestraintsareexpressedasthenon-linearequationsofstatewiththeinequalityrestraints,andthemodelprincipleofhybridsystemisintroduced.TheconclusionshowsthatitcomestruetoexactlycontrolpositionandattitudeoftheSRVbymeansofnon-linearmodelpredictivecontrol.Thetestinamodelbasinhasalsoprovedthattheabovemethodsareefficient.
简介:Uptonow,sometechnologyofneuralnetworksaredevelopedtosolvethenon-linearityofresearchedobjectsandtoimplementtheadaptivecontrolinmanyengineeringfields,andsomegoodresultswereachieved.Thoughitputssomequestionsovertodesignapplicationstructurewithneuralnetworks,itisreallyunknowableaboutthestudymechanismofthose.But,theimportanceofstudyratioiswidelyrealizedbymanyscientistsnow,andsomemethodsonthemodificationofthatareprovided.
简介:Thispaperdescribespathre-planningtechniquesandunderwaterobstacleavoidanceforunmannedsurfacevehicle(USV)basedonmulti-beamforwardlookingsonar(FLS).Near-optimalpathsinstaticanddynamicenvironmentswithunderwaterobstaclesarecomputedusinganumericalsolutionprocedurebasedonanAalgorithm.TheUSVismodeledwithacircularshapein2degreesoffreedom(surgeandyaw).Inthispaper,two-dimensional(2-D)underwaterobstacleavoidanceandtherobustreal-timepathre-planningtechniqueforactualUSVusingmulti-beamFLSaredeveloped.Ourreal-timepathre-planningalgorithmhasbeentestedtoregeneratetheoptimalpathforseveralupdatedframesinthefieldofviewofthesonarwithaproperupdatefrequencyoftheFLS.Theperformanceoftheproposedmethodwasverifiedthroughsimulations,andseaexperiments.Forsimulations,theUSVmodelcanavoidbothasinglestationaryobstacle,multiplestationaryobstaclesandmovingobstacleswiththenear-optimaltrajectorythatareperformedbothinthevehicleandtheworldreferenceframe.Forseaexperiments,theproposedmethodforanunderwaterobstacleavoidancesystemisimplementedwithaUSVtestplatform.TheactualUSVisautomaticallycontrolledandsucceededinitsreal-timeavoidanceagainstthestationaryunderseaobstacleinthefieldofviewoftheFLStogetherwiththeGlobalPositioningSystem(GPS)oftheUSV.