Contentslistsavailableat ScienceDirect Applied Ocean Research journal homepage: www.elsevier.com/locate/apor Adaptiverobustslidingmodecontrolofautonomousunderwatergliderwith inputconstraintsforpersistentvirtualmooring Hexiong Zhou a,b,c ,ZhaoyuWei a,c ,ZhengZeng a,c ,CaoyangYu a,c ,BaohengYao a,c, ⁎ ,LianLian a,c a State Key Laboratory of Ocean Engineering, Shanghai Jiao Tong University, Shanghai 200240, China b School of Naval Architecture Ocean & Civil Engineering, Shanghai Jiao Tong University, Shanghai 200240, China c School of Oceanography, Shanghai Jiao Tong University, Shanghai 200240, China ARTICLEINFO Keywords: Underwater glider Adaptiverobustslidingmodecontrol Virtual mooring Soft landing Motion tracking control Input constraints ABSTRACT Bottom-restingcapabilitywillenableanautonomousunderwaterglider(AUG)toperformlong-durationvirtual mooring.However,thetrackingcontrolofAUGsoftlandingontheseabedfacesanumberofchallenges.Itis crucialthatthecontrollerisabletoincorporatemanyimportantfactors,e.g.,modeluncertainty,environmental disturbancesandlimiteddynamicrangeofactuators.Thisstudy frst formulatesthesoftlandingtrackingcontrol problemofAUGforpersistentvirtualmooring.Anadaptiverobustcontrolschemebasedonslidingmodecontrol with a nonlinear disturbance observer is proposed that can efectively solve the challenges aroused by para- metricuncertaintyandunknowntime-varyingexternaldisturbances.Moreover,theefectofinputconstraintsof theAUGsystemisanalyzedforthe frsttimeandisdampedthroughadynamicauxiliarycompensator.Finally, simulation results demonstrate the superiority and inherent robustness of the proposed control scheme, in comparison with other techniques. 1. Introduction Autonomousunderwaterglider(AUG)isanunmannedobservation platform undergoing rapid development for oceanographic surveys. Oneofthemainglidermissions,whichisknownasthevirtualmooring, wherethevehicleperformsrepeatedproflestoacquiredataessentially continuous in depth at a predetermined location [1]. Recently, there hasbeengrowinginterestinextendingthedurationofvirtualmooring toundertakesciencemissionsoverlongerperiodswithoutsupervision, thereby reducing mission costs and extending their applicability. In particular,recentstudieshaveproposedbottom-restingtoimprovethe capability of the AUGs for persistent virtual mooring [2–4]. Bottom- restingallowsAUGtositandsleepperiodicallyontheseafoor.Dueto thenegligiblefowrateneartheseafoor,thevirtuallymooredunder- waterglidercanmaintainatafxedlocation forapredeterminedperiod with minimum power consumption. The capability of bottom-resting alsomakesitpossibletoexpandotherapplications,suchasunderwater wirelesspowertransferemployedontheseabed[5,6]andunderwater docking[7,8]. However, the control of AUG soft landing on the seabed faces a numberofchallengesinpractice.Unlikeglidinginasawtoothpattern wheretheAUGonlyneedstokeeptheattitudewithinanappropriate range to obtain an optimal gliding state [9], the aforementioned scenariohashigherperformancerequirementsforthecontrolprecision andrateofresponse.TheattitudeandvelocityoftheAUGmustmeet the permitted values timely for successful access to prevent damage fromcollisionswithseabed.Avarietyofmethodshavebeendeveloped and applied to the pitch and heading stabilization control problem of AUG.In [10],asystematicmethodbasedonlinearquadraticregularis proposed to control the vertical motion, and its advantage compared withconventionalproportional-integral-derivative(PID)isdetermined in [11],however,withaslowresponserate.In [12],atimesuspension approachisintegratedwiththemodelpredictivecontroltooperatethe AUGtofollowapre-defnedtrajectory,whichissuitable forlarge-scale ocean observations to track specifc phenomena. Active disturbance rejectioncontrolisappliedtopromotetheoperationalperformanceof theAUGin [13],whichcanreducetheovershootandsettlingtime.An integral sliding-mode controller is designed for a general type of the underwaterrobotin [14] andvalidatedtheefectivenessbycomparison withPDcontrolmethod.In [15],aninversesystemmethodispresented toaccomplishdecouplingforthenonlinearmotioncontroloftheAUG. However,inrealphysicalsystem,inputconstraints,modeluncertainty and unknown time-varying external disturbances pose serious chal- lenges to the motion tracking control of an AUG. Especially for the virtualmooring,theimpactofthesefactorscannotbeneglected. Input constraints are common problems in actual actuators of the https://doi.org/10.1016/j.apor.2019.102027 Received30June2019;Receivedinrevisedform28September2019;Accepted17December2019 ⁎ Corresponding author. E-mail address: yaobaoheng@sjtu.edu.cn (B.Yao). Applied Ocean Research 95 (2020) 102027 0141-1187/ © 2019 Elsevier Ltd. All rights reserved. T