中流的
1.Factors affecting flow classification in IP switch are analyzed. The n a fuzzy flow classifier (FC) is proposed. It can adjust its threshold accordin g to flows in the network and working state of IP switch.
2.Flow classification in IP switch is studied in this paperFirst,traditional X/Y flow classifiers are analyzed and factors affecting classification are concludedThen a new flow classification based on Hopfield neural network is proposed and how to select its parameters is discussedThe simulation results show that the Hopfield neural network(HNN)flow classifier can adjust its thresholds according to flows in the network and the number of VC is almost constantComparing to traditional X/Y classifier,HNN classifier can switch more packets with less VC
5.Because different fluidsaturation at different gas injection stage can leads to different streamline distribution,the two-phase model describes an ever-changing stream tubes.
6.Based on the needs of commercial operation, streaming media distribution in the ACR system must be manageable, operated and controlled. But the inherent openness of IP multicast technology make it very weak in the security management and commercial operation.
9.An improvement scheme about the Integrated oxidation ditch is obtained, for example big energy consumption and uneven flow velocity and so on.
10.The invention relates to apparatus for monitoring characteristics of a flow stream in a pipeline and in particular to dual chamber orifice fittings.
