简介:AnewapproachtogenerateAself-organizingfuzzyneuralnetworkmodel.Anonlinearcombiningforecastmethodbasedonfuzzyneuralnetwork.Anovelclustermethodinfrizzyneuralnetworks.AnovelrobustPIDcontrollerdesignbyfuzzyneuralnetwork.Arecurrentfuzzyneuralnetwork:learningandapplication.Astudyofchatterpredictioninendmillingprocess(fuzzyneuralnetworkmodelwithinputsofcuttingconditionsandsound.Aweightedfuzzyreasoninganditscorrespondingneuralnetwork.
简介:Anti-controlofchaosbasedonfuzzyneuralnetworksinversesystemmethod;ApplicationofArtificialIntelligenceTechniquesforClassificationandLocationofFaultsonThyristor-ControlledSeries-CompensatedLine;AutonomousNavigationinaKnownDynamicEnvironment;BETAFUZZYNEURALNETWORKAPPLICATIONINRECOGNITIONOFSPOKENISOLATEDARABICWORDS;Controlsystemusingfuzzifiedinputneuralnetwork;CoordinatedcontrolofEGRandVNTinturbochargeddieselenginebasedonintakeairmassobserver;DesignofIntelligentOptimalTrackingControlforRobotManipulator。
简介:AConcurrentFuzzy-NeuralNetworkApproachforDecisionSupportSystems,Adynamicallygeneratedfuzzyneuralnetworkanditsapplicationtotorsionalvibrationcontroloftandemcoldrollingmillspindles,Afuzzymodelingofactivemagneticbearingsystemandslidingmodecontrolwithrobusthyperplaneusinμ-synthesistheory……
简介:摘要:为了能够及时发现齿轮表面缺陷以及缺陷类型,我们提出了一种自适应分类框架,该框架根据任务需求调整分类粒度,并基于混合神经网络(AHNN)。AHNN结合了一维卷积和注意机制,增强了特征和通道之间的关系,并抑制了不敏感信息的影响,引入了个体特征选择方法,生成适合不同个体的特征子集,减小个体差异。实验结果表明,齿面磨损的细粒度和粗粒度分类的准确率分别为91.27%和96.31%,缺齿的细粒度和粗粒度分类的准确率分别为92.67%和97.28%,正常齿的细粒度和粗粒度分类的准确率分别为92.33%和96.39%。AHNN能够适应不同的分类粒度,降低个体差异,提高框架的通用性。
简介:[篇名]Aselftuningpredictivecontrollerbasedoninstantaneouslinearizationusingneuralnetworks,[篇名]Discrete-timeneuro-fuzzyadaptivecontrolbasedondynamicinversionforroboticmanipulators,[篇名]Fuzzyadaptiveoutputtrackingcontrolofaclassofcompositesystems,[篇名]Modelreferencefuzzyadaptivecontrolofdissolvedoxygenconcentration,[篇名]Model-referencefuzzyadaptivecontrolasaframeworkfornonlinearsystemcontrol,[篇名]Multivariablefuzzyadaptivecontrolofnonlinearsystems。
简介:落煤残存瓦斯量的确定是采掘工作面瓦斯涌出量预测的重要环节,它直接影响着采掘工作面瓦斯涌出量预测的精度,并与煤的变质程度、落煤粒度、原始瓦斯含量、暴露时间等影响因素呈非线性关系.人工神经网络具有表示任意非线性关系和学习的能力,是解决复杂非线性、不确定性和时变性问题的新思想和新方法.基于此,作者提出自适应神经网络的落煤残存瓦斯量预测模型,并结合不同矿井落煤残存瓦斯量的实际测定结果进行验证研究.结果表明,自适应调整权值的变步长BP神经网络模型预测精度高,收敛速度快;该预测模型的应用可为采掘工作面瓦斯涌出量的动态预测提供可靠的基础数据,为采掘工作面落煤残存瓦斯量的确定提出了一种全新的方法和思路.