Search results

Unfortunately the material data provided by manufacturers do not contain all necessary information to predict the behaviour of advanced composite structures using different analysis tools.
A significant reduction in calculations of the identification functional is achieved in this case in comparison with the conventional optimisation methods.
Plan of experiments Response surfaces Data from experiments FE analysis Error functional Identified parameters Fig. 1 Identification procedure.
In the third stage the numerical data obtained by the finite element method are used to determine a simple functions using the response surface method.

This paper will combine the theoretical study and practical data propose the method of designing the scale model for the research of the cathode protection .
The similarity theory In model experiments, only the models and prototypes remain similar ,we can get the results of a prototype according to the experimental data of the model. 
Design scale model of the ship cathode protection The scale model method is obtaining the similar mode scaled down the geometry of prototype according to the certain similarity criteria, complete the data measurement and analysis through the model simulation[6].
References [1] LIANG Cheng hao, YU Nan, WU Jian hua, Huang Nai bao, Research on physical scale model of ICCP system in double-zones of hull [J], Journal of Dalian University of Technology, 2008, Vol.48, No .5:656-660 [2] HUBBARD JC, BROOKS SH, TORRANCE BC Practical measures for reduction and management of the electromagnetic signatures of in-service surface ships and submarines UDT, [C] .1996: 64 ~ 65 [3] ZHENG Xian yan, Study on the Distribution of Protection Current and Protection Potential in Impressed Current Cathode Protection System Desing [J], Navigation of China, 2009, Vol.32, No.4 :92-95 [4] DEGIORGI VG, WIMMER SA, HOGAN E, et al.

Collaborative filtering is a technique that is complementary to learn, each other's interests, content-based filtering and user preferences, or the social behavior of the target data model predicted that the existing database user's preferences [4,5,6,7].
This means that they are facing performance issues, when the data is very large, sparse.
Input: the data source D = (U, I, R), the user - project focuses on matrix A Output: k a cluster Method: From the collection of items I = {Item1, Item2, ..., Itemn} n to retrieve all the items; From the basic user set U = {User1, User2, ..., User m} m to retrieve all users; Select from the m user score the most amount of users as the initial k cluster centers, denoted by {W1, W2, ..., Wk}; k-cluster C1, C2, ..., Ck are initialized to empty, denoted by set C = {C1C2, ..., Ck}; repeat for each user ui ∈ U for each cluster center Wj ∈ {W1, W2, ..., Wk} According to the formula (4) Calculate the similarity of ui and Wj sim (ui, Wj) end for sim (ui, Wm) = max {sim (ui, W1), ..., sim (ui, Wk)} Cluster Cm = Cm + ui end for The members of the cluster does not change until 3.
Kelleher, Experiments in sparsity reduction: Using clustering in collaborative recommenders, in Procs. of the Thirteenth Irish Conference on Artificial Intelligence and Cognitive Science, pp. 144–149.

It can be obtained by transforming formula (1) that, （2） Take a=αa, b= -αaαb, it can be obtained that a=0.31, b=0.38 by fitness analysis of test data in this paper and before in reference [6], so that αa=0.31, αb= -1.23.
Substitute them into formula (1), it can be obtained that, （3） Fig. 1 shows the scatters of all test data and the fitness line ( solid one ).
In this condition, the larger water to cement ratio brings benifit from reduction of cement content.
Based on the data in this test and before, the values of practical parameters in formula are obtained for calculating mixed strength of concrete.

According to the APICS (American Production and Inventory Management Association) definition of vendor-managed inventory: VMI is an optimized supply chain strategies, supply-side access to the customer's inventory data in this strategy, and is responsible for according to customer requirements to maintain its inventory levels. suppliers according to a predetermined cycle check the customer's physical inventory, the revocation of the inventory of damaged or expired, the completion of the inventory replenishment and continued to modify the inventory replenishment strategy ".
VMI mode of implementation of the conditions for the third party logistics (1) the parties must establish a state-of-the-art information systems Distributors via electronic data interchange (EDI) of the point-of-sale information to the suppliers, and suppliers via EDI delivery information sent to distributors, while using bar code technology and scanning technologies to ensure the accuracy of the data.
The implementation of VMI supplier master distributor sales information, inventory information, and many other confidential information, distributors must confirm that the information is not understood by competitors, and suppliers must also be recognized distributor inventory reduction as a result of such cooperation and increase effective space will not benefit competitors

The data obtained in the experiments was used to fit linear, logarithmic and second order regression models.
Within the experiemental temperature range an increase in reeling temperature leads to the reduction on the holding power of sericine hence allowing the reeling of silk to be done with ease.
Figure 2 The effect of reeling temperature on end breaks Effects of the variables on filament strength: The data obtained in the experiments was used to fit linear, logarithmic and second order regression models for the relations between filament strength and the two variables (reeling temperature and boiling time).
West, Multiple Regression as a general data analysis technique; Research in Science Education, 4(1) (1974) 182-192

This approach used historical data from earlier negotiations to extract the boundary of the agreement zone.
Using this data, it could combine both the advantages of the prediction of the opponent’s preference weight and the reservation price.
However, the approach lacks the ability to predict an agreement zone during the negotiating process which could change the course of learning process over time.The advantages of this technique are increment the chance of reaching an agreement and thereby reduction of the searching space’ssize.
Hence, we use theExponential Moving Average(EMA), a moving averagewhose weight is exponentially decreased over time, to extract the trend and to filter the noise out of the data.

We use this method to achieve the dimensionality reduction of the index system, and get the scientific index system, which provides a reference for the security of research about the aerospace launch mission.
According to deal with the data with SPSS software, the test results show as table 3: Table 3 KMO and Bartlett's test KMO .782 Bartlett's test Approximate chi-square 49.536 Df 25 Sig. .007 KMO statistic is the difference between the sum of squares of all variables simple correlation coefficient and the partial correlation coefficient, which is between 0 and 1.
(2) The factor analysis method bases on the provided data, analyze the correlation to different factors, then according to the different of correlation between correlate factors to combine different factors.
In this way, the correlation different, the combined way of factors is different, which is the results of subjective and objective data analysis and overcome the defects of some factors are combined subjectively in the evaluation process, so that the evaluation result are more reasonable, objective and scientific.

The EWMA controller for the SISO process is postulated as follows: (1) whereis the quality characteristic of the process output,is the process offset, andis the process gain, and,are the unknown parameters that must be estimated from the observed data.
In practice, the target value (T) and the receivable variance of the process can be set according to historical data.
This paper simulated over 200 different process data in the four different control strategies by using the R® program.
However, this reduction would consequently result in a high MSE value.

This approach based on conventional processing methods causes the process performance reduction and the problem of providing a high quality of the treated surface after cutting.
Analysis of the profilograms data shows that turning with the blanks advanced ultrasonic processing, Fig. 2b, 3b, is more preferable in comparison with the basic processing options, Fig. 2a, 3a.
The data analysis confirms the fact that advanced ultrasonic processing of blanks from thermoplastics allows to obtain subsequently a better surface layer of the part in comparison with the variant of the blanks base turning.
Thus the data analysis from Table 1 allow to conclude that these roughness parameters in the case of blanks with advanced ultrasonic treatment have a stable tendency to decrease.