简介:Inthenormaloperationcondition,aconventionalsquare-rootcubatureKalmanfilter(SRCKF)givessufficientlygoodestimationresults.However,ifthemeasurementsarenotreliable,theSRCKFmaygiveinaccurateresultsanddivergesbytime.ThisstudyintroducesanadaptiveSRCKFalgorithmwiththefiltergaincorrectionforthecaseofmeasurementmalfunctions.Byproposingaswitchingcriterion,anoptimalfilterisselectedfromtheadaptiveandconventionalSRCKFaccordingtothemeasurementquality.Asubsystemsoftfaultdetectionalgorithmisbuiltwiththefilterresidual.Utilizingaclearsubsystemfaultcoefficient,thefaultysubsystemisisolatedasaresultofthesystemreconstruction.Inordertoimprovetheperformanceofthemulti-sensorsystem,ahybridfusionalgorithmispresentedbasedontheadaptiveSRCKF.Thestateanderrorcovariancematrixarealsopredictedbythepriorifusionestimates,andareupdatedbythepredictedandestimatedinformationofsubsystems.Theproposedalgorithmswereappliedtothevesseldynamicpositioningsystemsimulation.TheywerecomparedwithnormalSRCKFandlocalestimationweightedfusionalgorithm.ThesimulationresultsshowthatthepresentedadaptiveSRCKFimprovestherobustnessofsubsystemfiltering,andthehybridfusionalgorithmhasthebetterperformance.Thesimulationverifiestheeffectivenessoftheproposedalgorithms.
简介:教育硕士在我国实行的时间不长,对于授课模式的探讨仍处于探讨阶段。目前的授课主要采用传统的方式,满堂灌,方法单一,课后教师对学生“不闻不问”,教学效果不好。如何改善目前现状,改变传统授课模式,提高教育硕士的学习效率成为最迫切的任务。笔者提供了几个建议:一是尊重学生个性,发展学生个性,而不是压抑个性;二是教师须治学严谨,每一堂课认真对待;三是采用分类教学,提高学生学习效率;四是采用实时网络远程教学,由于网络远程教学本身存在一些弊端。所以不能全部而只是有选择地使用;五是采用学分制。教育硕士大多数是在职人员,也有全日制的教育硕士,针对他们不同的情况,可以采用弹性学分制。