Application of Multivariate Statistical Analysis in Batch Processes

An Acoustic Emission (AE) signal processing and analysis system is developed for process and condition monitoring of Dense Medium Cyclones (DMCs), key units in coal preparation plants. This paper describes the use of a combination of statistical signal processing, multivariate statistical analysis and pattern classification methods to formulate robust soft-sensor predictive models for DMC process variables and wear state based on passive contact AE signals. Power spectral density and statistical moment ‘feature’ variables are extracted from AE signals. Principal Component Analysis (PCA) is used to evaluate the sensitivity of the AE ‘features’ to process variables and wear state. Principal Component Regression (PCR) and Partial Least Squares (PLS) are used to quantitatively model process variables. Hierarchical clustering is used to qualitatively model wear state and process variables. The AE hardware, signal processing and soft-sensor modelling system can be used for process and condition monitoring of other mineral processing equipment.

In today’s society, under the demand for high-quality talents, the professionals cultivated in colleges and universities should not only master professional skills but also need to have the spirit of innovation and entrepreneurial ability in professional aspects. In this paper, firstly, the multivariate statistical analysis method is explained, and the basic principles of the factor analysis model, principal component analysis method and systematic clustering method are given to analyze the data in the following article. Secondly, starting from the necessity of the integration of bi-initiative education and professional education of teacher training majors in colleges and universities, we constructed the 4344 specialized and innovative integrated talent cultivation model of teacher training majors and constructed an evaluation system to evaluate the effectiveness. Finally, based on the evaluation system, the data analysis of the specialized and creative integrated talent cultivation model was carried out using factor analysis, principal component analysis and systematic clustering method. The results show that in the factor analysis, the highest loading value of the first principal factor is 0.917, the contribution rate of the first principal factor is 39.67%, and the loading value of the principal component factor reaches the highest value of 0.925. The clustering analysis is based on the results of the factor analysis, and the respondents are divided into 4 clusters. The number of people in the 2nd category is more than 30 people, which accounts for about 31.41% of the total number of people. The method of multivariate statistical analysis can be used to analyze the data effectively for the specialized and integrated personnel training mode of teacher training in colleges and universities and also gives the path of specialized and integrated personnel training for teacher training.

Groundwater plays a dominant role in arid regions; it is among the most available water resources in Tunisia. Located in northwestern Tunisia, Oum Ali-Thelepte is a deep Miocene sedimentary aquifer, where groundwater is the most important source of water supply. The aim of the study is to investigate the hydrochemical processes leading to mineralization and to assess water quality with respect to agriculture and drinking for a better management of groundwater resources. To achieve such objectives, water analysis was carried out on 16 groundwater samples collected during January-February 2014. Stable isotopes and 26 hydrochemical parameters were examined. The interpretation of these analytical data showed that the concentrations of major and trace elements were within the permissible level for human use. The distribution of mineral processes in this aquifer was identified using conventional classification techniques, suggesting that the water facies gradually changes from Ca-HCO3 to Mg-SO4 type and are controlled by water-rock interaction. These results were endorsed using multivariate statistical methods such as principal component analysis and cluster analysis. The sustainability of groundwater for drinking and irrigation was assessed based on the water quality index (WQI) and on Wilcox and Richards's diagrams. This aquifer has been classified as "excellent water" serving good irrigation in the area. As for the stable isotope, the measurements showed that groundwater samples lay between global meteoric water line (GMWL) and LMWL; hence, this arrangement signifies that the recharge of the Oum Ali-Thelepte aquifer is ensured by rainwater infiltration through mountains in the border of the aquifer without evaporation effects.

The multivariate statistical analysis methods, using both multiple linear regression(MLR) and partial least square(PLS), have been applied to predict the reactor properties and the product quality of a direct esterification reactor for polyethylene terephthalate(PET) synthesis. On the basis of the set of data including the flow rate of water vapor, the flow rate of EG vapor, the concentration of acid end groups of a product and other operating conditions such as temperature, pressure, reaction times and feed monomer mole ratio, two multi-variable analysis methods have been applied. Their regression and prediction abilities also have been compared. The prediction results are critically compared with the actual plant data and the other mathematical model based results in reliability. This paper shows that PLS method approach can be used for the reasonably accurate prediction of a product quality of a direct esterification reactor in PET synthesis process.

This article applies principal component analysis (PCA) and cluster analysis to solve the customer classification problem of an actual power supply network. First, the article defines the index system which will influence the cost of power supply significantly. The index system includes power supply, maximum need, load rate etc. Second, the article uses PCA to filter the source data from these index items into several main components. Third, these main components will be analyzed with system clustering algorithm and k-means clustering algorithm to make clustering analysis and comparison and provide the calculation method about how to make the efficiency test of these clustering results. The clustering results prove that the source data can be well classified with PCA method with the new indexes reflecting the characteristics of the whole source data. At last the article also proves the similarity of the results obtained from system clustering and K-means clustering separately and verifies the two clustering algorithm all can be used to actual analysis. The multivariate statistical analysis proposed in this article can reflect the characteristic of customer's electrical behavior can be more objective and comprehensive that traditional methods and can a good tool for power price decision maker.

In this paper, the influence of economic variables on the annual electricity consumption in Beijing in 2000-2013 has been investigated. The purpose is to find out their relationship with single data sets and multivariate statistic methods. The multivariate methods used in this paper include ordinary least squares regression (OLSR), principal component regression (PCR), partial least squares regression (PLSR), and modified partial least squares regression (MPLSR). In comparison of these methods, the PLSR model is shown to perform better than the others. The effectiveness of the electricity consumption prediction models is illustrated and verified based on the practical data sets of electricity consumption in Beijing.

The applicability of multivariate statistical methods for the systematic analysis of results from the site‐directed experiments was tested. Dehalogenation data obtained with 15 single‐point mutants of haloalkane dehalogenase of Xanthobacter autotrophicus GJ10 has been related to various physico‐chemical properties of substituted amino acids. A two component model which explains 84% of activity data was derived. The most influential properties, required on position 172 in haloalkane dehalogenase were: aromaticity, main‐chain flexibility, refractivity and bulkiness, and were related mainly to the transition state/halide stabilization. The general usage of multivariate techniques for quantitative structure‐function relationships (QSFR) and protein engineering studies can be recommended. Score and loading plots from Principal Component Analysis (PCA) of physico‐chemical properties of amino acids can be used for planning and design of site‐directed mutagenesis experiments, while the models obtained from Projection to Latent Structures (PLS) can provide predictions on the effect of the mutations on protein functionality prior to its testing.

Prediction models based on multivariate statistical methods and their applications for predicting railway freight volume

The main objective of the present study is to perform risk assessment of groundwater contaminated by nitrate (NO3−) and evaluate the suitability of groundwater for domestic purposes in the Palani region of South India. Thirty groundwater samples were collected in the study area. Various groundwater quality analysis parameters such as the pH, electrical conductivity, total dissolved solids, total hardness, major cations (Ca2+, Mg2+, Na+, and K+), and major anions (Cl−, SO42−, F−, CO32−, and HCO3−) were adopted in this study to evaluate the drinking water suitability according to 2011 World Health Organization (WHO) standards. Piper and Gibbs’s diagrams for the tested groundwater indicated that, due to the influence of rock–water interactions, evaporation, and reverse ion exchange, the chemical composition of groundwater varied. According to water quality index (WQI) mapping results, 46.67% of the sample locations was identified as contaminated zones via GIS spatial analysis. Multivariate statistical analysis methods, such as principal component analysis, cluster analysis, and the Pearson correlation matrix, were applied to better understand the relationship between water quality parameters. The results demonstrated that 40% of the samples could be identified as highly affected zones in the study region due to a high nitrate concentration. The noncarcinogenic health risks among men, women, and children reached 40, 50, and 53%, respectively. The results illustrated that children and women occurred at a higher risk than did men in the study region. The major sources of contamination included discharge from households, uncovered septic tanks, leachate from waste dump sites, and excess utilization of fertilizers in the agricultural sector. Furthermore, using the nitrate health hazard integrated method with the conventional indexing approach ensures that groundwater reliability can be guaranteed, contamination can be explored, and appropriate remedial measures can be implemented.

In this study, we examined the metabolites from different parts of Acanthopanax senticosus and their role in alleviating damage caused by oxidative stress. We used UHPLC-QTOF-MS to analyze the chemical components in the root, seed, and leaf extracts of A. senticosus. Two multivariate statistical analysis methods—namely, principal component analysis and partial least square discriminant analysis—were used to distinguish the samples obtained from different parts of the plant. Using univariate statistics, 130 different metabolites were screened out. Among these, the relative content of flavonoids and terpenoids was found to be highest in the leaves, the lignin and phenolic acid content was highest in the roots, and the amino acid and phenolic acid levels were highest in seeds. An MTT assay was used to test the anti-H2O2 oxidative damage to PC12 cells in different parts of the sample. Lastly, using Pearson's correlation analysis, various metabolites from different parts of A. senticosus were correlated with their antioxidant effects from the corresponding parts. Fifty-two related different metabolites were found, of which 20 metabolites that were positively correlated to oxidative stress were present at a relatively higher level in the roots, whereas 32 metabolites that were negatively correlated were present at relatively higher levels in the seeds and leaves. The results of this study reveal the distribution characteristics and the antioxidant activity of different metabolites of A. senticosus and provide a reference for the rational development of its medicinal parts.

Application of multivariate statistical analyses to ItraxTM core scanner data for the identification of deep-marine sedimentary facies: A case study in the Galician Continental Margin

Based on the land surface temperature (LST), the land cover classification map, vegetation coverage, and surface evapotranspiration derived from EOS-MODIS satellite data, and by the use of GIS spatial analytic technique and multivariate statistical analysis method, the urban heat island (UHI) spatial distribution of the diurnal and seasonal variabilities and its driving forces are studied in Beijing city and surrounding areas in 2001. The relationships among UHI distribution and landcover categories, topographic factor, vegetation greenness, and surface evapotranspiration are analyzed. The results indicate that: (i) The significant UHI occur in Beijing city areas in the four seasons due to high heat capacity and multi-reflection of compression building, as well as with special topographic features of its three sides surrounded by mountains, especially in the summer. The UHI spatial distribution is corresponding with the urban geometry structure profile. The LST difference is approximately 4-6℃between Beijing city and suburb areas, comparatively is 8- 10℃between Beijing city area and outer suburb area in northwestern regions. (ii) The UHI distribution and intensity in daytime are different from nighttime in Beijing city area, the nighttime UHI is obvious. However, in the daytime, the significant UHI mainly appears in the summer, the autumn takes second place, and the UHI in the winter and the spring seem not obvious. The surface evapotranspiration in suburb areas is larger than that in urban areas in the summer, and high latent heat exchange is evident, which leads to LST difference between city area and suburb area. (iii) The reflection of surface landcover categories is sensitive to the UHI, the correlation between vegetation greenness and UHI shows obviously negative. The scatterplot shows that there is the negative correlation between NDVI and LST (R2 = 0.6481). The results demonstrate that the vegetation greenness is an important factor for reducing the UHI, and large-scale construction of greenbelts can considerably reduce the UHI effect.

Multivariate statistical analysis is often used in fault diagnosis. PCA is a commonly used multivariate statistical analysis method, based on the threshold method for fault diagnosis. But the PCA is the original space of the equivalent that didn't increase any amount of information, often resulting in a lot of minor failures difficult to diagnose. Deep Learning has a successful application in pattern recognition, linear representation of detail based on multi-level network. But Deep Learning is not explainable, only to give training results and there is no theoretical basis. On the basis of the original PCA, firstly expands and then reduces the dimension. The former is intend to fully express the information which cannot be expressed in original space by adding a projection, two projection,..., on the basis of original projection, and so on. And then generate a multi-level projection frame representing the space from the sample space high-dimensional feature. Experiment shows that the minor faults that cannot be detected in the original projection space can be detected under the projection frame of the new multi-level high-dimensional feature, and the detection rate of fault detection is improved.

The concept of life quality has been studied by specialists from a variety of scientific fields: economics, social geography, sociology, psychology, medicine, political sciences, and others. This contributes to the complementariness of the notion and broadens its interdisciplinary perspective, but on the other hand, it leads to a lack of unanimity in terms of the definition and measurement of the quality of life. Meanwhile, all developed countries in the world regard enhancing life quality as a priority of state policy. With the further advancement of our civilisation, quality of life will become a major issue in economic development. Therefore, monitoring this aspect of economic life, at both country and regional level, seems to be of particular significance. The paper aims to assess the suitability of selected methods of multivariate statistical analysis for the construction of a synthetic measure of objective quality of life. The study employs two methods of constructing synthetic measures of objective life quality: the linear ordering method – TOPSIS, and factor analysis. The results obtained by means of multivariate statistical analysis methods made it possible to create ratings of Polish and Belarusian regions in terms of objective quality of life and to further divide the regions into typological groups.

Multivariable Analysis Methods have been used widely in Spectroscopy Analysis. Partial least square (PLS) and principal component analysis (PCA) are the two most popular methods due to their excellent ability in data components analysis and results prediction. This work derived 1D/2D PLS and 1D/2D PCA based on the viewpoint of three-layer artificial neural networks, and uses theoretical proving to figure out the essences of these two methods. Two 2D experimental dataset was used to verify the calibration and prediction ability, furthermore, the similarity and dissimilarity of PLS and PCA. The finding showed that both the 1D/2D PLS and 1D/2D PCA methods use maximum covariance and minimum sum of square error to figure out the relationship between independent and dependent variables. The dissimilarity is that, weight vectors calibration of PLS is cross-correlation, and that of PCA is autocorrelation. The difference causes that the PCA method would keep more principal characters than PLS under insufficient sample among and provides better calibration ability. PLS would provide higher performance in prediction under sufficient sample among. For the needs of system implementation of spectroscopic measurement and analysis, this study designed "Multivariate Analysis Toolkits for LabVIEW" for the convenient implementation in automatic measurement system integration.