简介:ApplyingthegeneticprogrammingtomodelingofdiffusionprocessesbyusingtheCNNanditsapplicationstothesynchronization.Cellularneuralnetworkanditsapplicationinthediagnosisofabnormalautomobilesound.Cellularneuralnetworkinimagefiltrationtasks.Cellularneuralnetworksanditsapplicationforabnormaldetection-optimizationofthecellularneuralnetworksbydesigninganoutputfunction.
简介:Abiologicallyinspiredconnectionistsystemfornaturallanguageprocessing,Abiologicallymotivatedconnectionistsystemforpredictingthenextwordinnaturallanguagesentences,Acombinedmodelofwaveletandnenralnetworkforshorttermloadforecasting,Acomparativestudyofradialbasisfunctionneuralnetworksandwaveletneuralnetworksinclassificationofremotelysenseddata……
简介:Remarksonadaptiveneuralnetworkcontrollerusingreferencemodel,Revolutioncontrolofgeneratordieselenginebyneuralnetworkcontroller,Robustneuralnetworkcontrollerforvariableairflowvolumesystem,Selftuningneuralnetworkcontrollerforinductionmotordrives,Self-OrganizingNeural-BasedFuzzyControllerforTransientStabilityofMultimachinePowerSystemsUsingFlywheelBattery……
简介:Evolutionofaneuralnetworkforgaitanimation.Experimentalevaluationofanovelswitchcontrolschemetbranactivepowerlineconditioner.Fuzzylogicdecisionmechanismcombinedwithaneuro-controllerforfabrictensioninrobotizedsewingprocess,Hybridsteppingmotorpositionservosystemwithon-linetrainedfuzzyneuralnetworkcontroller.
简介:Aneuralnetworkcontrollerforsuppressionofwingrock;Aneuralnetworkpredictivecontrolsystemforpapermillwastewatertreatment;APIDneuralnetworkcontroller;Anartificialneuralnetworkapproachformotioncontrolofamagneticdiskdrivevoicecoilmotor;Aninternalmodelcontrolstrategyusingartificialneuralnetworksforaclassofnonlinearsystems;2002IEEEInternationalConferenceonSystems,ManandCybernetics(SMC02),vol.5:BridgingtheDigitalDivide-Cyber-development,HumanProgress,PeaceandProsperity;AutomaticGenerationControlforPowerSystemwithSMESbyUsingNeuralNetworkController;Chaoticsystemcontrolconsideringedgeofchaosusingneuralnetwork;DesignandimplementationofindustrialneuralnetworkcontrollerusingBackstepping;
简介:为了进一步优化神经网络算法,提高网络神经算法的速率并提高其稳定性,就现有BP算法所存在的收敛速度慢以及容易陷入局部极小值的弊病,我们将进一步通过一般改进算法解决在神经网络结构优化过程中依然无法解决的问题。依据遗传算法的特征,进一步在经过改进的压缩映射遗传的基础上提出了BP神经网络优化方案。泛函分析中压缩映射原理的应用,一方面解决了困扰人们的BP神经网络算法所固有的缺点,显著地提高了神经网络算法的收敛速度,而且解决了BP神经在运行的过程中和网络连接权值初值的取值紧密相连的缺点。经过大量的计算我们得到如下数据:经过优化改进后,训练时间节约了8.3%,训练步数降低了近17.4%。经过大量的研究实验表明:经过改进后的BP神经网络算法取得了良好的效果,十分具有应用价值。
简介:本文利用Lyapunov—Krasovskii稳定性定理和线性矩阵不等式技术,得到了多时滞区间神经网路全局鲁棒稳定性的一个新的稳定性规则。该规则推广了最近文献中的一些结果,并通过数值仿真证实了结果的正确性。