Search results

Abstract: Linear Discriminant Analysis (LDA) [1] is a well-known method for face recognition in feature extraction and dimension reduction.
LDA extracts features of the original data in a way which selects the most discriminable features between classes.
The main difference between the 2DLDA and LDA algorithm is the representation of data[4].
It has been demonstrated that the feature vectors produced by this method could classify better. 1 Preliminary Knowledge:2DLDA The goal of 2DLDA is to extract features that can well discriminate a set of data belonging to a number of classes.
This part is as same as that of 2DLDA, except . 3 Experimental result and analysis I test my method on ORL[7]face data pool.

Evidently from the data represented in Table 1 that grain refinement is rather modest despite the reasonable deformation level accumulated.
From the data represented one can conclude that the temperature of SPD processing play a key role in grain refinement in molybdenum.
Literature data reveal the same structural features in chromium and tungsten processed by different ways of SPD.
Let consider the thermal stability of the structure of Mo processed by HPT and MF (the data for ECAP-Mo are not available yet).
The similar data are obtained in [3].

Illustrative size reduction data corresponding to ultrasonic frequencies of 20 kHz, 500 W are shown in Fig. 2(b).
Based on turbidity data, Fig. 4(a) illustrates the noticeable difference between the two decantation levels of the 30-minutes sono-fragmented sample, using 20 kHz, 500 W and 58 kHz, 500 W, respectively prior to start of blending.
Figs. 4(b) and 4(c) illustrate that good blending is obtained after 30 minutes, based on turbidity and LPC data, respectively, for the 20 kHz sono-fragmented particle sample followed by 132 kHz sono-blending.
After completion of several experiments using different combinations of fragmentation and blending frequencies (e.g., 20/33, 20/58, & 20/132), optimum time data obtained from turbidity measurements have been plotted as a function of ultrasonic frequency in Fig. 4(d).
By analyzing sono-fragmented and sono-blended alumina ceramic particles, which are a stratified mixture of nano/ sub-micron particles, optimum blending time was determined on the basis of turbidimetric data.

So, if the period of immersion increases one has a reduction in D and Da.
It is also obvious from the above data that the TMCs are generally superior to the CPJ control mortar in the resistance to Cl- ion penetration.
A linear relationship in the form of y = ax + b seems to be best fit data R2 value of more than 0.95.
Mix Flexural strength (MPa) 2 days 7 days 28 days 56 days 90 days 120 days TM0 4.53 (0.58) 4.94 (0.64) 7.7 8.2 (1.06) 8.54 (1.11) 8.94 (1.16) TM2 4.38 (0.58) 4.62 (0.61) 7.5 8.12 (1.08) 8.24 (1.10) 8.43 (1.12) TM4 3.8 (0.52) 4.35 (0.60) 7.23 7.91 (1.09) 8.04 (1.11) 8.18 (1.13) TM6 3.83 (0.53) 4.24 (0.60) 7.19 7.5 (1.04) 8.01 (1.11) 8.34 (1.16) (*) The data in parentheses are ratios of 28-day mechanical strength (Flexural).
(f) About 23.5% reductions in leached depth in NH4Cl solution.

Data reduction strategies compensating for closure and Kmax effects on crack growth and design curves will be presented.
Long fatigue crack growth curves for all studied alloys tested at R=0.1:(a) DKapp data and (b) DKeffdata after adjusted compliance ratio (ACR) [3] closure corrections.
Stress Ratio, Initial Crack Size,and Kmax Effects on Fatigue Crack Growth Data and Design Curves.In Figs.6(a,b), stress ratio,Kmax, and microstructurally small crack effects are shown for the cast A535-F (450μm) alloy.
(a,b) Stress ratio, Kmax, and small crack effects; (c,d) data normalization and prediction.
The data were normalized to incorporate Kmax effects, Fig. 6(c), and design curve predictions were made in Fig. 6(d).

The findings of this study revealed that the immersion of natural rubber in d-limonene led to a significant reduction of tensile properties due to chain scission and bond breaking as well as cross-linking within the rubber.
Additionally, the samples exhibited noticeable dimensional changes with greater concentrations of d-limonene leading to more substantial swelling and ultimately causing a reduction in tensile strength.
The data obtained from the specimens were presented as part of this study's results as shown in Figures 1 and 3.
The data from the specimens are presented in Figure 2 and 4.
This can weaken the intermolecular forces and interactions within the rubber matrix, causing a reduction in tensile strength.

Application of Chemical Storage in Smart Grid and Analysis on Loss Reduction Chunyan Kong1, a, Shenhang Yu2, b and Ying Sun2,c 1Department of Chemistry, Dezhou University, Dezhou, China 2School of Electrical Engineering, Shandong University, Jinan, China aemail: chunyankong@163.com, bemail : yushenhang@163.com, cemail : sy@sdu.edu.cn Keywords: Chemical storage; Flow battery; Smart grid; Energy saving Abstract.
Energy Saving Analysis This paper uses the run data of a feeder within 24h in a city distribution network as an example, simulates and analyzes the energy saving effect after the chemical storage batteries connected to the grid.
With the example of 24h operating data, we can found that the total value of the feeder load in a day showed significant bimodal characteristic, and load peaks appeared at 11:00 and 18:00, so we set the scheduling policy: charging the battery from 22:00 to 6:00 of the next day, discharging the battery in two stages, from 8:00 to 12:00 and 17:00 to 21:00, and the battery storage does not exchange with the system to obtain the best smoothing effect.
The data of the example shows that a smoother curve corresponds to a lower loss rate.
Above the perspective of the stable power supply, CO2 emission reduction and energy saving, research and application of the new efficient storage technology have become imperative, while the chemical storage is a efficient storage technology which mainly used in the terminal network and the supporting application with the solar power, wind power and other renewable energy generation.

Generally, the following simulation setting data (see Table 2) have to be inputted to the software for the rotating cases.
Table 2 The simulation setting data inputs for FEM simulation Forming Machine Hydraulic Press Simulation Module 3D Rotating Speed for the Lower Die 0.2rad/sec Constant Shear Friction at the Upper/Bottom interface 0.3/0.1 Reduction Ratio 18.67~32.67% Moving Speed for the Upper Die 2mm/sec Flow Stress of Upper-Layer Cylinder SUS304H : Flow Stress of Bottom-Layer Cylinder SAE1039 : Geometry of Upper-layer Cylinder Geometry of Bottom-layer Cylinder Fig. 3 presents the corresponding results including the effective stress, the effective strain and material flow of the double-layer clad cylinder after compression whether considering the rotating effect.
Results and Discussions The compression forces obtained by the Slab analysis and FEM simulation under the given reduction ratios for various clad height ratios can be compared whether considering the rotating effect.
Fig. 4 is variation of compression force in non-rotating compression with reduction under different clad height ratios.
And due to the embedding effect which hard layer embeds into the soft layer during the compression, the soft layer size will bigger with the increasing of reduction ratio.

Pujana applied ultrasonic vibration on drilling Ti6Al4V and observed force reduction and temperature increment during the process [2].
Including all the benefits that ultrasonic brings to drilling process, thrust force reduction is the most important one.
Many researchers have discussed the thrust force reduction through both of the theoretical analysis and the experiments.
A data acquisition card recorded the data and sent it to the computer.
For the cutting lips, the reduction of thrust force is more significant than that of the chisel edge.

In this paper, multi-embedded systems distributed and cooperative control program takes instead of multi L2 server mode, one L2 data collection and record server was remained.
To seek for the reduction resources under the condition of costs equivalence, its mechanical properties and thermal conductivity are better than that of single stainless steels.
Table 1 Functional division of the two systems System 1 System 2 Work roll speed Work roll diameter Backup roll diameter Opening Plate thickness Plate width Material Strain rate Strain degree Deformation temperature Coefficient of friction Plate temperature Initialization data and process data are provided by OPC server, centralized computation is transformed into distributive computation, and thus the overall system performance is improved.
The conclusion is that the error is less than 6% with the measured data comparing with the calculation results (Shown as Table 2).
Table 2 The experimental results Rolling Passes Entry Thickness Reduction (mm) Temperature () Rolling Force(KN) Error (%) Calculation Actually 1 203 25 1237 6190.62 6001.03 3 2 178 25 1227 6470.41 6534.45 1 3 152 20 1216 5949.85 6178.64 5 4 132 25 1201 8428.42 8179.18 3 5 107 15 1194 6121.74 6312.19 4 6 91 15 1174 7205.72 7201.23 0 7 76 13 1165 7114.23 7112.33 0 8 64 13 1134 8741.51 8579.28 2 9 51 8 1122 6465.86 6667.81 3 10 43 8 1078 8178.58 8179.18 0 11 36 8 1064 10161.97 9823.9 4 Acknowledgment This work was financially supported by National Youth Science Foundation of China (51205269) and “973 Program” (2011CB612204).