Search results

Compared with the method of reduction roasting under resistance heating, the activation energy of microwave reduction roasting was small, and the reaction rate was fast.
AB Yu [4] studied the process of microwave carbothermic reduction of stannic oxide and discussed the influencing factors of the reduction rate.
Specific experimental processes are as follows: crude niobium concentrate powers were mixed with a certain percentage of carbon evenly in the mortar, pressed into pieces, then put into porcelain crucible with thermal insulation set, and then heated in microwave oven until reaching the enactment temperature and staying there for a while, at the same time, the electron balance was used to record the weight loss data and the temperature were recorded by person.
According to Eq.7, the TGA data of crude niobium concentrate were calculated, which were obtained from the reduction roasting process at the temperature range of 300°C~850°C, and the result was shown in fig.10.
Judging from the correlation coefficient R by the regression, it can be seen that the kinetic parameter values obtained by non-isothermal analysis were closer to the data obtained in the experiment.

In order to lower the measurement system errors in aluminum electrolysis reduction, and find out the abnormity of process parameters, this paper considered the process parameters in the production process, used mean-range control chart in statistical process control to analyse aluminium level data during aluminium electrolysis reduction, designed and implemented the statistical process control system for aluminum electrolysis reduction data, presented process improvement method for aluminium level data.
If the data points is less, we take n=1.
Daily report, effect report, laboratory data, the measurement data, etc. all have cell number field and date field.
The system selects and analyses the data of 4045# experiment pot's aluminium level measurement data, which have two months' data.
There is no abnormal data.

It was found that the presence of aquatic worms in the reactor significantly improved the sludge reduction by increasing the total suspended solids (TSS) reduction (Table 1).
Accordingly, sludge reduction was 36–77% [6].
Table 1 Literature data of sludge reduction with worms in aerobic wastewater treatment processes Operation conditions Main results Controlsb References Pilot activated sludge system; 20 oC; domestic wastewater Ya=0.14 SVI ↓ Control Y=0.22 Wei et al. (2003) [3] TSS↓ 39-65% VSS↓ 0.5-6.3mg/mg worm/d Y=0.10-0.27 SVI↓ Control deducted in calculations Control Y=0.25-0.49 Liang et al. (2006)[7] TSS ↓ 48(±45)% No control Wei & Liu (2006)[8] VSS↓ 0.2-0.8mg/mg worm/d Effluent TP↑ Control deducted in calculations Huang et al. (2007)[9] Pilot activated sludge system; 18-23 oC; domestic wastewater COD↓ 18-67% Y=0.15 SVI↓ phosphate↑ Control COD↓ 20% Control Y=0.4 Rensink & Rulkens (1997)[10] Sludge disposal or TSS↓ 25-50% SVI↓ Control Ratsak (1996)[4] TSS↓ 36-77% The hydraulic load↑5-15% Hendrickx et al.(2009)[6] aY refers to the sludge yield (kgSS/kgCODremoved) bControls refer to the conventional biofilter (without worms) Although the presence of worms in the aerobic wastewater
It was found that the presence of earthworms in vermifilter significantly improved the sludge reduction, the BOD5 reduction, the COD reduction and the excess sludge stabilization was also enhanced by reducing the ratio of volatile suspended solids to suspended solids (VSS/SS).
Sludge reduction with a novel combined worm-reactor.

If we can extract this information from the database, then we will create a lot of potential profit and value for the owner of the data, and this kind of technology of mining information from large database is data mining. 1.1 Theoretical basis of data mining The basic theory of data mining can be attributed to the following several aspects: 1) Data reduction: according to this theory, the basis of the data mining is to reduce the description of the data.
In a large database, data reduction can translate fast approximate response to queries.
Data reduction technology mainly includes the singular value decomposition behind the main component analysis of drive elements, wavelet, regression log-linear model, histogram, cluster, sampling, and the index tree structure. 2) Data compression: according to this theory, the basis of data mining is to compress the given data.
For example, model found can be seen as a form of data reduction and data compression.
It mainly includes the following main components: Visual user interface Model knowledge assessment Data mining engine Database or data warehouse server Data cleaning Data integration Data filtering Database Data warehouse Knowledge base Figure1.

The XRD data indicate the composite consists of elementary Fe and Cu deposited on the inert EG.
Chemical reduction processes have become a new focus in recent studies.
But nitrate reduction by Fe0, unlike halogenated hydrocarbon reduction, is relatively sensitive to solution pH.
The data acquisition was performed using Chromeleon 6.5 SP2 (Dionex, USA).
The XRD data (not shown) of the composite consists of the elementary Fe and Cu deposited on the EG.

The simulated data showed equilibrium of ilmenite with precipitated phases formed at different temperature.
The initial stage of ilmenite reduction was fast because of iron oxide reduction by CO [7].
XRD data on 1000-1100°C are presented in Sharifah Aishah et. al. study [5].
The X was higher than the reduction at 5 wt. % that had reduction of 39.8%.
After 180 minutes of reduction, samples with 10 wt. % FeCl3 had a reduction of 69.0%.

The sample data after dimension reduction is used for LSSVM training, and the LSSVM classification model is obtained, which forms the ability of detecting unknown intrusion.
Then, the sample data after dimension reduction is inputted into the LSSVM classification model for training, and the classification model is used for intrusion detection.
Step4: Using the reserved features (attributes) after attribute reduction and the original sample data, construct the simplified samples.
Therefore, the 41 features of the original data are simplified to only 17 features after attribute reduction, and most of the irrelevant features are eliminated.
The attribute reduction method of RS can remove the irrelevant or unimportant input features, and reduce the dimension of the input data.

In this paper, coal-based reduction on flotation middling from iron ore containing carbonate at donganshan was studied, during which the effect of reduction temperature, reduction time, C/O mole ratio and feed layer thickness on reduction process were carried out.
It is a good reductant for coal-based reduction.
From physical chemistry of metallurgy view point, the increase of reduction temperature can promote iron reduction reaction.
Effect of Reduction Time.
The data reveals that metallization degree grew slowly from 1.0 to 1.5, but begins to take off at 1.5, while further increase in C/O results in slightly change.

Many approaches for dimension reduction exist in the field, e.g. by feature selection using some objective criteria [18, 8,11].
In the following, we will use only the good data (we will refer to them as 'the data').
What is the shape of the multidimensional data cluster containing data points representing subsequent data vectors, 2.
It is obvious that dimension reduction works in favor for further processing.
Low dimensional visualization using PCA and self-associative neural network Principal component analysis (PCA) PCA is a well established method used for reduction and low-dimensional visualization of data [16,17,4].

SSDP makes full use of both labeled and unlabeled data to construct the weight incorporating the neighborhood information of data.
The labeled data points are used to maximize the separability between different classes and the unlabeled data points are used to estimate the intrinsic geometric structure of the data.
This leads ones to consider the dimensionality reduction methods to represent the data in a lower dimensional space.
It performs dimensionality reduction by projecting the original high dimensional data into the low dimensional linear subspace spanned by the leading eigenvectors of the data’s covariance matrix.
In many practical applications of pattern classification and data mining, one often faces a lack of sufficient labeled data, since labeling data often requires expensive human labor.