简介:Inordertostudytheeffectsofwetcompressiononatransoniccompressor,afull3-Dsteadynumericalsimulationwascarriedoutundervaryingconditions.Differentinjectedwaterflowratesanddropletdiameterswereconsidered.Theeffectofwetcompressionontheshock,separatedflow,pressureratio,andefficiencywasinvestigated.Additionally,theeffectofwetcompressiononthetipclearancewhenthecompressorrunsinthenear-stallandstallsituationswasemphasized.Analysisoftheresultsshowsthattherangeofstableoperationisextended,andthatthepressureratioandinletairflowratearealsoincreasedatthenear-stallpoint.Inaddition,itseemsthatthereisanoptimumsizeofthedropletdiameter.
简介:提出了一个全新的动态公交客流预测模型,使用智能卡数据和网络爬取的常州市兴趣点(POI)数据、天气记录数据,建立一种特征融合的长短期记忆递归神经网络(Feature-fusionLSTMRNN)模型,并应用于常州市地面公交客流集散点的智慧卡(SmartCard)刷卡客流预测.将以上模型用于刷卡客流预测实验,结果表明经过良好数据训练后的特征融合的长短期记忆递归神经网络可以在刷卡客流预测中保持较高的稳定性、准确性和泛化能力.