Search results

Therefore, the computed displacement data provided by FEA codes must be translated into forms acceptable by optical software.
With the displacement data and the corresponding code, the optical performance of the mirror can be analyzed.
In the validation process, the free-free mode condition was performed to obtain measured data and then used for comparison with data predicted by the FEA model.
The experimental data were obtained via hammer testing using a Photon Plus signal analyzer that was connected to a PCB tri-axial piezoelectric accelerometer (Model 356B21).
The measurements were carried out by impacting the mirror at point 1 and acquiring the response data at points 1 to 36.

Survey data shows that, for daily life, the ideal indoor noise level shall not be higher than 40dB(A), and 55dB(A) or higher noise level may bring about complaints; ideal noise level for sleeping is 30dB(A) or less.
Moreover, low sound reduction is found within the range of medium frequency.
Its measured sound reduction-frequency characteristic curve is shown in Fig. 1.
Fig. 2 GRC sound insulation - frequency characteristic curve In addition to field measurement results, the sound reduction testing data of lightweight boards produced by some China-based manufacturers has failed to measure up to sound insulation standards.
Furthermore,the measured weighted sound reduction (Rw) of 100mm-thick fiber-reinforced partition board from a certain company limited of new building materials with a surface density of 42.5 kg/m2 is 37.5dB; its sound insulation - frequency curve is shown in Fig. 3.It is observed from above-noted test data that the weighted sound reduction of single-layer walls with one-tier structure made from single material is generally lower than 40dB. 3.

High-level media data decoding depends on the lower-level media data, so media data at low level should have higher priority. [11] pointed out that the request priority mechanism of scarcity data block ensure the diversity of the data block in overlay network, which is conducive to large-scale exchange of data between nodes.
In order to enable the process of data block distribution faster and smoother in the whole overlay network, data block priority should include playback of the data block cut-off time, the scarcity of data blocks, block level where the media are included.
Data requests can carry the current playback time, so that the data provider can determine whether the data block (Chunk) urgency, or whether it has failed.
Gradient-based model of data distribution strategy.
Further observed that the improvement of broadcast continuous is mainly due to the significant reduction of the buffer time, because the gradient model based scheduling policy is conducive to an emergency request for the data block to get more upload bandwidth, in this way to avoid emergency required data is delayed.

However, many of these methods stop at the theoretical stage, because they lack the breadth and depth of data required in practice.
The equation parameters were obtained through data analysis.
Therefore, the results may have problems such as duplicated data or leakage.
According to the characteristics of water environmental degradation and data available for the Taihu Basin, the protection costs method and classification method were selected to estimate the water environmental degradation costs.
Because of data limitations, some evaluation methods failed to operate in practice.

Remove stop words can reduce the dimension of feature space and noise reduction.
Text representation The text is a string consisting of a large number of characters, belonging to the non structured data, can not be directly used for cluster analysis.
Text data not only has very high dimension but also has some characteristics of directional data, using the cosine similarity measure for clustering results significantly better than used in text clustering.
Provide DF, IDF feature selection method. 2) Clustering module After processing resources preprocessing module, feature vector can be obtained for the clustering ensemble, the data into text clustering module.
Corpus data for classification of news corpus (corpus collected in accordance with the SOGOU-T Webpage format), classification system, collection of data statistics.

Introduction A hard disk drive industry has been more competitive for supporting the needs of data storage that increasing exponentially.
Therefore, this paper focuses on defective reduction in an in-house recycle process of a hard disk drive media.
The company data in Fig. 1 on November 2012 shows that an in-house recycle process built defective rate related to media about 23%.
Based on data from 30 days and 1,000 samples each day, P chart in Fig. 11 shows the average of defective rate of sunray defect is 3.203% and all is in the specification.
Senjuntichai: Spiral Defect Reduction of Hard Disk Disk Media, Applied Mechanics and Materials, Vol. 421 (2013), p. 93-98.

Observed data were utilized to obtain various performance parameters such as brake power, brake specific fuel consumption, brake thermal efficiency at various loads and throttle settings.
The collected test data are used to calculate brake power, brake specific fuel consumption and brake thermal efficiency.
Observed data in this table indicate that specified rated output of this diesel engine lies between best power & best economy zone at wide throttle opening (for the maximum duration of injection set in existing injection timing of the engine).
Minimum bsfc and maximum output for both the fuels reduces with the reduction in throttle (% reduction in duration of injection).
Percentage reduction in maximum output with the use of blend from their values with diesel reduces with reduction of throttle (duration of fuel injection).

Accidental factor error is data abnormal due to human error, needs to reject or correction error in data preprocessing process.
Basic steps of Reverse Engineering are data measurement, data preprocessing, data segmentation and surface reconstruction, CAD mold building [5].
Fig. 1 Trimming process of test Taking member stretching static test as an example, trimming the error in data preprocessing process based on reverse engineering, so as to obtain more precise data.
According to the analysis of data error in Pre-experiment and data preprocessing, feedback test scheme to get more precise test results.
Data reduction methods for reverse engineering [J].

This simulation system results in a reduction of cost because several experiments on all possible combinations of input data are no longer needed [11,12].
The available data was divided into 70% for the implementation of the training process and the remaining 30% was used as a test data for evaluating the trained model.
The values of RMSE are 0.4745 and 0.6847 for training and testing data respectively.
Moreover, the outputs of test data confirm the strength of this model in predicting the results.
Where the root mean square error (RMSE) values for both training data and test data were 0.474 and 0.684 respectively.

Data: Three types of data are used in this study.
Firstly, published data.
All of the published data can be downloaded from the internet.
Secondly, survey data.
Finally, calculation data.