Search results

In [9], neuro-fuzzy classifier is used for qualitative and quantitative classification of the data which was pre-processed by PCA algorithm.
Good class separability in both training and verification data was obtained by dimension reduction.
In this figure, the unfilled markers show training data.
In each case, the same LDA transformation matrix developed in the classification process of the training data was applied to feature vectors of the 25 verification experiments (5 of each TG).
A successful classification of the five TGs, for both training and verification data, was achieved in a reduced feature space according to Fig. 3.

Welded structures with the use of 1420 alloy [2] ensured the weigh reduction down to 24% as compared to the conventional structure of pressurized sections assembled of commercial D16 (2024) semiproducts with the use of mechanical riveting and sealants.
Such loads give the typical mark of the current crack front position on the test results in order to obtain FCGR data.
The maximum, minimum and average values are given according to the data 8.
It should be noted, that the data about the fatigue resistance, obtained for the hand and die forging are characterized by the increased statistic scatter, which indicates to the insufficient stability of fatigue characteristics for this semiproducts.
The assembly for EBW Conclusion The obtained data show the possibilities of the 1420 alloy hand and die forgings and EBW potentialities of 1420 alloy, in particular, for the production of large-size products even with the account of possible fatigue life decrease the expense of the scale effect, residual stresses after welding of a fullscale structure, technological scattering.

(15) (4) If and iteration number , then return step (3). 3 Simulation Results and Analysis 3.1 Data Pre-Processing There are many factors have influence on HTPV (), which including regional foreign-related tourism income (), the number of regional foreign-related tourism days (), regional GDP (), regional per-capita disposable income (), regional year-end general population (), regional highway traffic mileage (), highway traveling passenger turnover volume (), regional tourism general income () and domestic travel rate of regional urban residents ().
The data we used in this paper is from the city of Changsha, as shown in Table 1 [1, 3].
From Eq.16, normalization processing of data in Table 1 is made for the simulation
The data from 1997 to 2003 is used as training samples, and the data from 2004 to 2006 is used as test samples.
Liao, Aluminum reduction cell situation’s forecast based on wavelet neural networks, Control Engineering of China. 19 (2012) 290-293

Whatever the index selection, data acquisition, or evaluation method, on the basis of the science should.
Establish the DSM index system is a very big project, index workload when the more, collect data takes the manpower and material resources, also can give data processing trouble, index system is not the bigger the better, so set up index system, we must earnestly screening, to avoid duplication, to fully consider the quantification and data index has difficulty and reliability, try to select those representative comprehensive index and the main index.
This paper first consult literature regarding the large power demand side management aspect content, and then press usually index system of the principles, methods, formulated some experts investigation views and large amounts of data, text material, again according to the index system analysis theory established the index system (table 1).
Table 1 Comprehensive Evaluation Index System DSM on Railway Temporary Power System Railway temporary power system demand side management comprehensive evaluation index system Power side costs Can avoid peak jose capacity costs Can avoid fuel costs Can avoid turbine abnormal start-stop costs Can avoid the compensation costs caused environmental pollution Power grid side investment Reduce sales electricity income Input promo fees benefit Can avoid grid investment costs Can avoid power operation and maintenance costs Power supply reliability Customer satisfaction load ratio User side investment Initial equipment investment total costs Increase operation and maintenance costs benefit The average investment can be recovered Avoid compulsory limiting power costs social Environmental benefits Reduce carbon emissions Reduce sulfur dioxide emissions Reduce nitric oxide emissions Reduce emissions of solid waste The wastewater reduction Development

In order to increase laser processing efficiency and quality, acquiring some shape data of the work piece is essential and these robot systems can be more and more robust.
The laser hardening process is shown in figure 2, part surface point cloud data is obtained by 3D measurement system, and then the part model is reconstructed from the point cloud data in reverse engineering and imported into the path planner to do path planning and generating robot programs, as well as simulating and optimizing instructions in ROBOGUIDE, in the end, program is downloaded to robot controller, accomplishing hardening process.
According to FANUC robot language format, a conversion program based on Mathematica6.0 is developed which can automatically convert the data into robot motion commands.
Fig. 3 Flow chart of measurement system start Point cloud of parts Data reduction Triangular grid Segmentation Three-dimensional model Three-dimensional digital scanning Path planning & programming Planning Principle.

Preliminary experiments involving a new method of a controlled melting stop at the silicon surface during flash lamp processing [4] demonstrate a drastic reduction of undulation on the wafer during such processing.
This is not in accordance with the observed TEM data, which shows up to 90% of the upper SiC layer (54 nm) can be dissolved. 0 50 100 150 200 250 300 350 14 16 18 20 22 24 Tim e (m s) Melt depth (nm) SOL Substrate 0 50 100 150 200 250 300 350 75 80 85 90 95 100 Pulse Intensity (J/cm 2) Melt depth (nm) SOL Substrate However, carbon can easily diffuse through the liquid silicon as the diffusion coefficient is large.
Data given in [6] gives the diffusion co-efficient D to be 4.9×10-5 cm2/s at the melting point in liquid silicon.
This shows that these results are consistent with the observed TEM data, and that carbon transport is possible despite the low solubility limit.
The results are consistent with experimental data and we are able to use the data to optimize the layers and perform virtual experiments with a new melt stop technique on silicon surfaces.

Introduction Radio Frequency Identification (RFID) is a non-contact automatic identification technology featuring high precision, long reading distance and large storage data volume, thus has been widely applied in logistics, car industry and medical field.
In the RFID system working process, i.e., data communication between reader and label, data can be easily interfered by electromagnetic or destroyed by noise, resulting in data communication error.
Thus before demodulation of the received radio-frequency signal, interference and noises should be removed in order enhance the reliability of RFID system work and the accuracy of data transmission.
Research of Noise Reduction Algorithm Based on Real Time[J].

Determination of the fatigue strength is very expensive and require a lot of time.Usually, roll manufacture catalogs contain data about chemical composition of the material, the roll hardness on the surface and hardness drop from the surface, tensile and bending strength of the core and the picture of metallographic structure of the material.
Roll basic data The broken rolls of the 3-high-roughing mill had dimension Æ450 x 1400 mm and they made in steelworks "VALJI d.o.o", Slovenia.
These rolls are nodular iron cast rolls with the pearlitic base and data are shown in the tables 2 and 3 and Fig.4.
Fig. 9 Fatigue strength (S-N curves) for the axial and bending tests of the roll material Conclusion Tensile strength (325¸425 MPa) and bending strength (500¸700 MPa) of the roll material were known as well as hardness (350¸420 HB), see Table 3, but crucial material data for the fatigue life prediction – fatigue strength, was missing.
Obtained fatigue test data should be used for the rolls fatigue failures analysis as well as improvement of the rolls and rolling procedure in the future.

National Lightning Detection Network™ and applications of cloud to ground lightning data by electric power utilities, IEEE Trans.
Rakov, Lightning characteristics based on data from the Austrian lightning location system, IEEE Trans.
Yokoyama, Lightning occurrence data observed with lightning location systems in Japan: 1992-1995, IEEE Trans.
Bevington, Data Reduction and Error Analysis for the Physical Sciences, New York, 1969
Underwood, 2005: A Method to Identify the Optimal Areal Unit for NLDN Cloud-to-Ground Lightning Flash Data Analysis.

The results show that: CODCr removal rate and decolorization rate of landfill leachate reduce with the reduction of the influent pH, the increment of water concentration and the extension of the reaction time; Under the optimized operating conditions: the influent pH of 8.10, the influent CODCr of 16611 mg/L, the reaction time of 90 min, the CODCr removal rate and decolorization rate of landfill leachate reach 56.8% and 83.7%, respectively.
The experimental data are disposed, and the results of these experiments were shown in Fig. 1 (a)—1 (e) and Table 1.