Search results

One-hot encoding is used to convert IP, categorical to numerical data.
IP being a nominal categorical factor, one-hot coding is necessary to convert categorical data to numerical data from Gyroid, Line, and Tri-hexagon to 1, 2, and 3 respectively.
This error represents how well the model fits the training data.
This reflects the model’s predictive performance on new data.
Plot distribution of (a) training, (b) testing, and (c) checking data vs prediction of FIS.

The obtained flow stress-strain data was used to develop the Arrhenius constitutive model of which material constants considered the compensation of strain.
Toward this end, isothermal hot compression tests were conducted and the stress−strain data were then employed to derive Arrhenius constitutive equations.
The slight variation in the slope of the lines can be attributed to scattering in the experimental data points.
The Arrhenius predicted data are not so satisfactory under the 1000ºC temperature condition, because the microstructure of Ti−6Al−4V alloy changes from α phase to β phase in this temperature zone.
In most regions of the forging, the effective stress is between 10 and 40 MPa, which is agreed with isothermal compression test data.

Carbon Dioxide Geological Storage: A New Choice to Reduce Carbon Dioxide Emission in China Xufeng LI1, 2,a, Fenghua Shi3,b , Xiaofeng Jia1 and Gengbiao Qiu1 1 Center for Hydrogeology and Environmental Geology, China Geological Survey, Baoding, China, 071051 2 School of Water Resources and environment, China University of Geosciences (Beijing), Beijing, China, 100083 3College of Urban and Rural Construction, Agricultural University of Hebei, Baoding, China, 071051 affslxf@163.com, bshifenghua@hebau.edu.cn, Keywords: Climate Change; Global Warming; CO2 Emission Reduction; CO2 Geological Storage; China Abstract.
And the total amount of CO2 emission from the major industrial sources (power plants cement production steelworks refineries ethylene ammonia ethylene oxide and hydrogen) in China is estimated to be 29.6315×108 t with the production data of the above various plants [15].

Exposure time per dataset was 15 min and the CCD-detector was used for automated data acquisition.
A fit of the data from 1 to 39 hours with a diffusion limited precipitation model [7] results in the solid curve shown in Fig.4.
The model function represents the data well up to 33 hours, i.e. to the end of the first holding step.
Such behavior is found qualitatively in our data.
Fig.3: Part of the data (intensity) from Fig.2 in the range from 2 to 10 hours.

At present, the common measuring method includes Fuzzy Comprehensive Evaluation, Analytic Hierarchy Process (AHP), Data Envelopment Analysis method, and Neural Network (NN), etc.
This index reflects the GM data management situation of the enterprise, and it is difficult to quantify. (6) Anticipation to social environmental projects.
The data are gathered from the Internet, including the corporate portal website, related news and reports [9,10,11].
And the standardized data are shown in Table 3.
Because the input data are between [0, 1], the transfer function for the hidden layer is Logsig, and for the output layer is Purelin.

Using the model to indentify the natural frequency of bearing rotor under different parameters, then compare identification value with experimental values shows that projections in good agreement with the experimental data.
SVM achieved data reduction.
In order to give full play to the good characteristics of SVM, we must choose specific data sets for SVM parameters.
Concrete data are shown in Table.1and Table.2: Table1: Standard input 1 2 3 4 5 6 7 8 9 2.16 2.16 2.16 2.16 2.16 2.16 2.16 2.16 2.16 7800 7800 7800 7800 7800 7800 7800 7800 7800 0.02126 0.05345 0.06029 0.09242 0.11319 0.13273 0.15754 0.17357 0.18916 0.03792 0.08874 0.09864 0.13681 0.15642 0.18511 0.21039 0.24223 0.2911 0.03991 0.05533 0.07633 0.08488 0.11204 0.13079 0.14621 0.16141 0.19165 0.2424 0.53322 0.7971 0.9644 1.121 1.3247 1.5558 1.6024 1.9018 0.22476 0.31964 0.4526 0.63948 0.84893 1.0029 1.0961 1.2341 1.3611 0.2581 0.41908 0.76388 0.95073 1.0367 1.3355 1.5853 1.7033 1.8386 20 20 20 20 20 20 20 20 20 Table2: Standard output 1 2 3 4 5 6 7 8 9 4.0957 3.7861 3.3013 3.4227 2.4099 2.4684 1.9517 1.6133 1.6367 65.4413 37.172 25.02 21.35 14.646 14.347 11.205 8.8576 9.3318 Meanwhile, in order to meet the needs of support vector machine model and have a better performance , this paper normalized the experimental data ,so that the data fall in [0,1].
Concrete data are shown in Table.3: Table 3: Error comparison of the analyzed results NO.

Time Series Analysis Time series is defined as the time order of collection of observations is associated according to the time changes and data sequence [3].
Table 1 Biogas generator temperature for five days times biogas generator temperature Monday Wednesday Friday Saturday Sunday 1 57.73 58.19 57.96 57.63 57.73 2 58.84 57.76 57.98 55.94 58.84 3 57.25 58.28 58.36 56.35 57.25 4 57.69 56.59 57.97 56.56 57.69 5 56.89 57.89 57.45 57.14 56.89 6 58.78 58.36 57.12 57.58 58.78 7 59.62 58.98 56.69 58.35 59.62 8 58.53 59.34 57.67 58.96 58.53 9 57.15 58.25 57.99 59.19 57.15 10 59.16 58.37 58.69 59.67 59.16 11 58.24 59.24 59.15 58.78 58.24 12 58.96 58.35 58.56 58.10 57.67 Use the former 59 data and time series analysis methods to predict the 60th data, and then compare forecast data with actual data to determine time series analysis methods for reliability indexes of the rolling forecast voltage effectiveness.
Dynamic data processing theory and methods - time series analysis [M].
Dynamic data system strategy time series modeling [J].
Dynamic data processing theory and methods - time series analysis [M].

Bagga et al. [16] created a matching between predicting data of tool wear and manual measurement data.
The data is processed in the hidden layer mathematically, then transferred to the output layer with final weight.
To implement ANN, Weka (a type of data mining software written in java language) was used.
This research focuses on classify tab; which provides several machine learning algorithms for the classification data.
Besides, a percentage split of 70% was used to divide data into two groups: 70% of data was used for training (56 samples) and 34% of data used for testing (24 samples).

The real time monitoring data characterize the different spatial and time domains is the information carrier of structural health.
Therefore, the data acquisition and processing system are the bridge connected to monitoring systems hardware and software systems[6-8].
Besides, it should consider the time interval of data acquisition, data standardization, measurement uncertainty, purification of data and other problems.
Data analysis and processing is that extract the structural performance and damage state characteristic factor from the signals of acquisition system[9].
On-line temperature monitoring systems for aluminum reduction based on fiber bragg grating [J].

Objectives The overall objective of this study was to develop and adapt existing computational methodologies to produce an engineering tool that is capable of predicting the effects of rocket plumes on ejection seat aerodynamic characteristics.The developed methodology utilizes a rocket nozzle and plume computational method coupled with an existing computational environment for ejection seat flows.Specific objectives of this study were:1/development of boundary condition modules for controllable propulsion.2/implementation of local grid refinement for improved plume resolution,3/validation of methodologies against 2D and 3D rocket plume data,and 4/perform a demonstration calculation for 4th Generation Ejection Seat with rocket power.
Two sets of basic validation data were chosen.The first data set was taken from a 0.036 scale B-1 escape capsule test[9].The data set consisted of calibration measurements of model plume shapes for the escape capsule rockets at different jet to free-stream pressure ratios.Excellent agreement between the predicted and measured plume shapes was obtained for all three pressure ratios.The second set of experimental data was chosen from the works of Fearn and Roth.
Comparison of Measured and Computed Nozzle Thrust Profiles In Fig. 6,the thrust versus pintle position profiles predicted by the series of CFD solutions is compared to the profiles derived from the Aerojet test data.The equilibrium and frozen chemistry simulations predict almost identical thrust profiles,and both compare well with the test data.Both the CFD and test results produce a nearly linearly thrust profiles as a function of pintle position.At the lower thrust settings,the CFD simulations tend to over predict the nozzle thrust.This phenomenon was also reported in the Aerojet analysis[13]and was attributed to thermal expansion and ablation of the nozzle throat.
Computational Grid at Symmetry Plane for Multi-Domain, Many-to-One Gridof Upper Rocket Geometry Including Exit.reduction in the normal force coefficients when the rockets are on.These predictions are consistent with previous wind tunnel test findings[15]The reason for the increase in the force coefficients is due to the lower pressure zones created at the back of the seat by the inward direct rocket plumes.In Fig. 10,the pressure contours on the surface of the seat and at the symmetry plane for the multi-domain grid for both the rockets-off and rockets-on test cases is presented.
Acknowledgments The authors of this report would like to thank Dr.Anatha Krishnan of CFERC for his support in adapting CFD-ACE to perform equilibrium simulations of ejection seat rocket plumes,and Mr.Matt Thomas for providing information on the composition and properties of solid propellants.The authors would also like to thank Mr.Joe Morris and Mr.Bill Barnette of Aerojet for their full cooperation in providing detailed geometry and test data on the PEPS system.