Search results

Introduction Along with more and more extensive application of the Internet , the control and management of the Internet is also put forward new demands, such as the implement in the network security measures and implementing quality of service(QoS)control.These are need for network monitoring and implementation of certain operational status of network measurements to measure the status of network operation which can be measured in differentnetwork parameters, but in a variety of measurement methods, the sampling is the most commonly used, which rely on sampling some of the data to estimate the present state of the network[1].However, in the actual process of sampling, there will be leakage sampling situation, for example, a fixed sampling algorithm in N is a message selected 1 message, to flow collection.
Flow measurement as meet specific attribute network message in a clustering method, which not only can satisfy the fine-grained network management needs in the data storage and data processing and also has a better advantage.
g) Network overview Analysis of the flow of information which is stored in the server to the reduction of the corresponding network at the sampling rate.
The flow sampling method based on the application group, firstly identifing the received packet and getting the packet corresponding to the type of application; according to the type of application to the message is sent to the corresponding application packet sample space, and record the number of packets is sent to sample space of the application group; the number of packets reaches a preset sampling value, determines the message for the sampled message; analyzing the sampled message flow information acquisition, and the flow of information is sent to the server, so the server stores the flow information ( five yuan ), network overview analysis flow information stored in the server., to the sampling rate for the reduction of the corresponding network.The sampling method for the received message identification, and identification of messages are sent to the corresponding application packet sample space, then, for each individual sample space sampling.

Omitted data, analysis, and related discussions can be found in [24].The elastic modulus has been evaluated adopting the ultrasonic time of flight technique, according to the following equation: Vl=E1-νρ1+ν1-2ν (1) being Vl the ultrasonic longitudinal speed, ρ the density and ν the Poisson ratio of the processed material.
The Poisson coefficient has been assumed, considering reference data, as 0.33.
The application of the conventional contour method is based on four consecutive steps: specimen cut, acquisition of the relaxed surface, data reduction, and stress computation.
Experimental data relative to both halves of the cut specimen (Fig. 1) have then been combined and processed in the MATLAB environment, in order to fit the experimental data to a smoothing surface.
The following data refer to four test cases, as indicated in Table 1.

The virtual city 3D model needs a lot of data.
Because only graphic data can be stored in AutoCAD,the method pays attention to separating graphics data and the graphic datain the data collection.
Set up texture lab Acquisition and processing of texture data.There are two main types of texture data acquisition method.
Graphics processing operations mainly includes: distortion correction, removing the mask, image enhancement, and data redundancy reduction.
Read data via automatic programming, batch processing of the model and reduce the workload and the amount of data that scene.

As a result of the presence of noise in the measured vibration signal has a great influence on the results of calculation of fractal dimension, Therefore the empirical mode decomposition method for noise reduction of gear vibration signal is used, calculation fractal dimension, extraction fault feature of Gear in different conditions.
The decomposition step of signal as follows: 1) Determine the local extreme point of all the signal, then use the three spline curve method to connect all local maxima and form upper envelope; 2) then use the three spline curve method to connect all local maxima and form lower envelope, the upper envelope and lower envelope should envelope all data points; 3)the upper envelope and lower envelope is signed m1(t), The original signal x(t) minus m1(t) is the first component h1(t): (1) 4)If h1(t) don’t fulfill these necessities, Then take h1(t) as the original data, Repeat steps 1)~3), Get the average m11(t), Judge whether the component h11(t)=h1(t)－m11(t) meet the conditions of IMF.
If it doesn’t fulfill, then cycle K times. 5)Let separated from (2) Take as the original data, repeat steps 1)~4),get the second component to satisfy the IMF condition, repeat n times, obtained some satisfying IMF condition component of the signal.

The paper presents a method for identify fault section in distribution network, which uses the traveling wave data recorded at the substation only.
This paper presents a new single-end fault section estimation method in distribution system using wavelets, which just utilizes the traveling wave data recorded at substation to identify the fault section.
The sample frequency is up to 10MHz, and every simulation will reserve sampling data in 4ms period which contains 1ms data before fault and 3ms post-fault data.
Conclusion The paper presents a single-end method to identify the fault section in distribution network, which uses the traveling wave data recorded at the substation only.
Matsushima, Development of a new fault locator using the one-terminal voltage and current data, IEEE Trans.

It is well-known that seismic disaster will cause serious damage, so the prediction and evaluation of seismic loss before earthquake event happened can provide foundation of disaster reduction program.
Macroscopic Vulnerability Model Based on GA-ANN 2.1 Collection and Pre-processing of Sample data From ref.4 -5, we picked out 80 sample data about the loss estimation results of earthquake which happened at different seismic intensity zones.
Normalizing the input-output data sets between certain ranges is an important step, this improves input-output generalization to yield accurate forecasts.
In this paper, eq.2 is used to normalize the raw data, after such pre-processing, the convergence time is significantly shortened
The data about seismic intensity, population density, per captia GDP and unit area GDP were input to the well-trained ANN model, the results as shown in tab.3 showed the output of the loss ratio of different areas, thus the estimation result could obtained.

Conventional GM (1, 1) model for long-term data to predict accuracy is not high, but metabolism GM (1, 1) model can be made up for conventional GM (1, 1) model of this one defect [1, 2].
Consider continuous differential equation, we can get the cloud forecasting model of system, (3) The solution of this differential equation is: (4) So the reduction solution (discrete form) of the single factor system cloud gray SCGM (1, 1) model is: (5) In this paper, we predicted the stress and camber of bridge by establishing the SCGM (1, 1) model with equal dimension and new information.
Set a SCGM (1, 1) model, after accumulating x(0)and we can get this sequence: X(1)= (0.868, 1.840, 2.792, 3.732, 4.719), then we can get the reduction solution R(0)(k)=30.709709×(1-e-0.029434)× e0.029434(k-1), R(0)=( 0.891, 0.917, 0.945, 0.973, 1.002, 1.032), so predict stress value is Z(6) = R(0)(k)Y(6) =1.032×10.34=9.79MPa, The result error is e=-4.406%.
For control the elevation of the box girder top surface in the construction process, we take the ratio of the measured data and theoretical camber value as raw data sequence in building model, to predict the camber of next section.
In order to ensure the accuracy of the measurement data, we should pay attention to measuring method and the selection of measurement instruments in construction control.

Application of Pt modified Al diffusion coating to CMSX-4 single crystals resulted in fatigue life reduction at 500 °C and in fatigue life increase at 900 °C [4].
More thorough data on fatigue behavior of the untreated material are reported elsewhere [7,8].
Experimental data were fitted by the Manson-Coffin law log 2Nf = (1/c) log εap - (1/c) log fε,
Experimental data are fitted by the Basquin law Fig. 1.
Experimental data were approximated by the power law log σa = log K´ + n´ log εap

An increased value of h corresponds to a reduction in emitter width or doping, and reduces the amount of stored charge near that junction.
As well as comparison with experimental data, finite element simulations have been used to validate physical models used in the diode electro-thermal model as well as to investigate dynamic device behavior.
It can be seen that the simulation output closely matches experimental data, though a discrepancy remains in that the final current tail is not captured by the model.
Finite element simulation results have indicated that this is due to the reduction in carrier drift velocity after most of the charge is removed from the base region, resulting in a longer transit time and thus the current tail characteristic at turn-off.
This model has resulted in improved matching of device power losses whilst reducing the amount of empirical data employed for simulations, thus facilitating efficient and accurate prediction of device power losses in power electronics circuit applications.

Till now, some meaningful studies have been made on the modeling of wind turbine, the DFIG system can be simplified in order to save computational time or to eliminate hard to obtain data.
It is also shown that the reduction/simplification of the model of converter and induction machine does not notably influence DFIG transient response.
The accuracy of the generic models will be studied by comparing simulation results with measurement data provided by manufacturers.
Assessment of DFIG simplified model parameters using field test data[C] .Power Electronics and Machines in Wind Applications (Chicago,America ,December16-18,2012),pp.1-7
[6]M.signh,K.Faria.Validation and Analysis of Wind Power Plant Models Using Short-Circuit Field Measurement Data[C].Power Energy Society General Meeting IEEE, (Detroit,America ,July24-29,2009),pp.1-6