82 results about "Multivariate statistics" patented technology

Thousands of process and equipment measurements are gathered by the modern digital process control systems that are deployed in refineries and chemical plants. Several years of these data are historized in databases for analysis and reporting. These databases can be mined for the data patterns that occur during normal operation and those patterns used to determine when the process is behaving abnormally. These normal operating patterns are represented by sets of models. These models include simple engineering equations, which express known relationships that should be true during normal operations and multivariate statistical models based on a variation of principle component analysis. Equipment and process problems can be detected by comparing the data gathered on a minute by minute basis to predictions from these models of normal operation. The deviation between the expected pattern in the process operating data and the actual data pattern are interpreted by fuzzy Petri nets to determine the normality of the process operations. This is then used to help the operator localize and diagnose the root cause of the problem.

A system and method for monitoring a process in a process plant and detecting an abnormal condition includes collecting data representative of the operation of the process, performing a multivariate statistical analysis to represent the operation of the process in a known state based on a set of collected reference data, where the reference data includes a statistical measure of the operation of the process in the known state. The system and method may further include representing the operation of the process in an unknown state based on a set of monitored data, where the monitored data includes a statistical measure of the operation of the process in an unknown state, and using the output of the multivariate statistical analysis as an input, and comparing the process in the unknown state to the multivariate statistical representation of the operation of the process in the known state to determine the operational state of the process.

This invention pertains generally to the field of multivariate statistics, and in particular to new methods for the analysis (e.g., chemometrics) of chemical, biochemical, and biological data, including, for example, spectral data, including but not limited to nuclear magnetic resonance (NMR) spectral data. These methods are useful, for example, in metabonomics, proteomics, transcriptomics, genomics, etc., and form a part of other methods, for example, methods for the identification of chemical species, methods for the identification of biomarkers that are useful in methods of classification, diagnosis, prognosis, etc.

The invention discloses a method for screening malignant ovarian tumor markers according to blood serum metabolic profiling. In the method, the Ultra Performance Liquid Chromatography-mass spectrometry technology is employed to analyze blood serum to obtain blood serum metabolic profiling; then the multivariate statistics method is employed to analyze data concerning malignant ovarian tumor and blood serum metabolic profiling of healthy people, so as to screen maker spectrums. The screening method of the invention features good repeatability and good forecasting property of the screened markers. Proven by discriminant classification analysis, the screening method enjoys an average forecasting precision rate of 90.90% and a positive relevance ratio of 82.65%.The maker spectrum covering multiple compounds can be obtained in a single analysis, which indicates that the marker is suitable for high flux analysis and enjoys the prospect of being promoted to large-scale sample screening and clinical application.

The invention discloses a small-fault detection method and device based on multiple moving average. The method includes the steps that sample data under normal working conditions are collected, a PCA model is built through the PCA method, and a load matrix P is acquired; multivariate statistics SPE and T2 of a first sliding time window are acquired based on the sample data of process variables under normal working conditions of the first sliding time window at each sampling moment; first statistics characteristics of the SPE and T2 of the first sliding time window are extracted; multiple sliding average treatment is carried out in terms of first statistics characteristics of the SPE and T2 of a second sliding time window, and second statistics of the SPE and T2 are acquired; a fault judgment interval for small fault detection is determined, and fault detection rules are defined; sample data of process variables of a working site are collected, the second statistics characteristics of the multivariate statistics SPE and T2 are acquired in the working site according to the load matrix P, and whether small faults occur or not is judged according to the fault detection rules.

A system and method of monitoring and diagnosing on-line multivariate process variable data in a process plant, where the multivariate process data comprises a plurality of process variables each having a plurality of observations, includes collecting on-line process data from a process control system within the process plant when the process is on-line, where the collected on-line process data comprises a plurality of observations of a plurality of process variables and where the plurality of observations of the set of collected process data comprises a first data space having a plurality of dimensions, performing a multivariate statistical analysis to represent the operation of the process based on a set of collected on-line process data comprising a measure of the operation of the process when the process is on-line within a second data space having fewer dimensions than the first data space, performing a univariate analysis to represent the operation of the process as a multivariate projection of the on-line process data by a univariate variable for each of the process variables, where the univariate variable unifies the process variables, and generating a visualization comprising a first plot of a result generated by the multivariate statistical representation of the operation of the process and a second plot of a result generated by the univariate representation of the operation of the process.

The invention discloses a serial bone plate design method based on an average thighbone model. The design method includes the steps of 1 building the average thighbone model, wherein parameters are defined, a constraint relation of the parameters is set up, and a multivariate statistics mathematic model is set up; 2 parameterizing characteristics of a bone plate, wherein the characteristics of the bone plate are generated, the characteristic parameters of the bone plate are defined in a layered mode, and the characteristics of the bone plate are compiled progressively; 3 designing the bone plate serially, wherein a mapping relation between the characteristic parameters of the bone plate and curved surface morphological parameters of the average thighbone model is set up, and serial bone plate design is achieved by modifying the parameters. A scientific basis is provided for reasonable and serial bone plate design in the field of computer-assisted orthopedic operations, and the serial bone plate design method is of great significance in improving the bone plate design quality and shortening design time.

A method and apparatus for detecting faults. A set of data samples is received, the set of data samples including multiple process variables. One or more multivariate statistical models are adapted, wherein adapting includes applying a change to at least one univariate statistic of the one or more multivariate statistical models if the change is greater than a threshold value. The one or more multivariate statistical models are used to analyze subsequent process data to detect faults.

This invention pertains generally to the field of multivariate statistics, and in particular to new methods for the analysis (e.g., chemometrics) of chemical, biochemical, and biological data, including, for example, spectral data, including but not limited to nuclear magnetic resonance (NMR) spectral data. These methods are useful, for example, in metabonomics, proteomics, transcriptomics, genomics, etc., and form a part of other methods, for example, methods for the identification of chemical species, methods for the identification of biomarkers that are useful in methods of classification, diagnosis, prognosis, etc.

The invention discloses a method for studying metabolic difference of transgenic rice and non-transgenic rice. The method analyzes rice seed extract with liquid chromatography-mass spectrometry (LC-MS) technology to obtain metabolic profiling of rice, studies integral difference of metabolic profiling data between transgenic rice and non-transgenic rice by multivariate statistics method, and finds out compound with significant contribution to metabolic phenotype difference, to provide a basis for evaluation of unintended effect and safety of transgenic rice. The simple and rapid analysis method with good repeatability is suitable for batch analysis of practical samples.

The invention belongs to the environment field of agricultural resources, and is an evaluation method for intensivism use of land, comprising, proposing that the intensivism use includes 4 parts of input intensity, utilization degree, output effect and social benefit of land, establishing a evaluation index system and an index determination method of intensivism use of land and proposing the evaluation index and a determination method thereof for intensivism use potential of land. The invention determines a relatively scientific and rational evaluation indicator system by 3-grade screening procedure of frequency analysis, Delphi and multivariate statistics for a lot of primary evaluation indexes and the weight of each index by using the combined method of principal component analysis and hierarchical analysis. The invention improves the single index screening and weight determination to analyze the dynamic change of intensivism use for land and determine the potential at a certain time point, and is a rare quantitative evaluation method for the field of evaluation for regional agricultural resources or quantitative calculation for intensivism use potential of land.

The invention discloses a soft measurement method for detecting the temperature distribution of billet inside a metallurgy heating furnace. The method selects a plurality of key variables affecting the temperature distribution of billet inside the furnace to form a process detection variable set, applies partial least square technology based on multivariate statistics projection principle to establish a soft measurement model between the temperature distribution variable of billet of the metallurgy heating furnace and the process detection variable and determines the optimal pivot numbers through cross check technology. The method avoids complicated process mechanism analysis, has convenient field implementation and high soft measurement precision and is particularly applicable to such industrial occasions similar to the temperature distribution detection inside the metallurgy heating furnace with high dimension and very rich process data. The soft measurement method can be used for monitoring in real time or directing actual production, improve the product quality, increase the yield and realize the energy conservation and consumption reduction of the metallurgy heating furnace overall.

The invention provides a humidity and temperature monitoring-based method for detecting moisture of grains at internal points of a granary. The method comprises the following steps: building a relational model between the equilibrium moisture and the temperature and the humidity of the grains at the internal points of the granary, determining initial values of all parameters in a formula by adopting a static weighing method and multivariate statistics regression; building a database of parameters of different grain varieties after model parameter correction; acquiring the corresponding temperature and humidity values of the internal points of the grains based on a system for detecting the humidity and the temperature of the internal points of the granary; acquiring the corresponding temperature and relative humidity value of single point or multiple points inside the granary through data acquisition, and computing by directly calling the rapid moisture detection formula and formula parameter values corresponding to different grains by a system, and therefore, the moisture of the single point or the multiple points inside the granary can be rapidly detected.

The invention relates to a method for determining a time for retraining a data mining model, comprising the steps of: calculating multivariate statistics of a training model during a training phase; storing the multivariate statistics in the data mining model; monitoring at least one distribution parameter of application data as a function of time during a deployment phase; and evaluating reliability of the data mining model based on the multivariate statistics and at least one distribution parameter.

The invention relates to a precoding design method of a maximized minimum signal to noise ratio in a large-scale MIMO (multiple input multiple output) system. At first, according to an instant receiving signal to noise ratio of a sub-channel in every radio frequency port of a ZF (zero frequency) receiver used by a base station terminal in an uplink, a mean receiving signal to noise ratio is obtained by using a multivariate statistics method; the sub-channel is optimized on the basis of the maximized minimum mean receiving signal to noise ratio rule; according to the independence of distribution type MIMO ports, the optimization of the mean receiving signal to noise ratio is decomposed to be a precoding matrix design under the limit of independent power in ports and the total power restraint power distribution optimization design between ports; finally, the optimal precoding matrix is obtained. The precoding design method of the maximized minimum signal to noise ratio in the large-scale MIMO obtains the optimal porecoding matrix by bysing the statistical information of a channel only, and has low system feedback cost; meanwhile, in comparison to the traditional power distribution method, the method can obviously improve the mean symbol error rate performance of a system, and thereby promoting the feasibility of the method in actual application.

The invention provides a method and a device for determining history matching adjustment parameters in numerical reservoir simulation. The method comprises the following steps of determining a function relation of oil production and water yield of an oil well and a water injection rate of a connected well layer according to reservoir data in the numerical reservoir simulation; carrying out multiple linear regression analysis on the function relation of the oil production and the water yield of the target oil well and the water injection rate of the connected well layer, and determining a multiple linear regression formula of the oil production of the target oil well and the water injection rate of the connected well layer; and determining the to-be-adjusted history matching parameters of the target oil well according to regression coefficients of the determined multiple linear regression formula of the oil production of the target oil well and the water injection rate of the connected well layer. By adopting a method of multivariate statistics, relation is established between the water absorbing capacity of each layer of a water injection well and the yield of a production well, which are obtained through analog computation, and through analyzing the influence of each water absorbing layer on the yield of the oil well, the layer position is judged and adjusted, the parameters of the corresponding small layer are adjusted and modified, and the history matching is guided, so that the purpose of detailed history matching is achieved.

The invention relates to a system for acquiring discontinuous emission spectra data, whereby the data is analyzed by a multivariate statistic model or equivalent model, such as principal component analysis and the use of the system for flow cytometry.

If a threshold discrimination is performed with variable Z=0 using discriminant analysis, that is useless unless know-how is accumulated through visual judgment and actual operation. A discriminant function is computed using a plurality of parameters which make pass / fail judgment factors and the results of that pass / fail judgment. With respect to the discriminant function, a histogram is generated for pass category and for fail category. Then, a threshold is determined based on the standard deviation in the individual categories so that an intended rate of flowout and rate of overcontrol will be obtained. The acceptability of pass / fail judgment objects is judged based on the threshold. Thus, the rate of flowout and the rate of overcontrol can be controlled as intended. Further, high-performance pass / fail judgment can be implemented without accumulating know-how.

The present invention relates, e.g., to a method for generating and analyzing multi-factorial biological response profiles, comprising (a) exposing each member of a plurality of expression control sequences, each of which is operatively linked to a heterologous reporter sequence, independently, to at least about three stimuli from a first set of stimuli, wherein at least about two of the stimuli in said first set of stimuli are, optionally, combined in an intra-set combinatorial fashion; (b) detecting a first category of responses of said expression control sequences to said stimuli; and (c) generating a response profile for each of said expression control sequences. The method may further comprise (d) exposing each of said members of the plurality of expression control sequences, independently, to one or more additional sets of stimuli, optionally wherein at least about two of the stimuli in each of said additional sets of stimuli are combined in an intra-set combinatorial fashion, in an inter-set combinatorial fashion with set first set of stimuli; (e) detecting the first category of responses of said expression control sequences to the stimuli in d); and (f) generating a response profile for each of said expression control sequences, which includes the responses detected in b) and in e). Raw response profiles are preferably analyzed by multivariate statistical methods, using a computer.

A rapid and efficient method for novel biological substance screening by surface analysis has been developed using Time-of-Flight Secondary Ion Mass Spectrometry (ToF-SIMS). This method relies on the surface screening of an array of micro-organisms grown on porous membranes, which had previously been in contact with a solid growth medium. TOF-SIMS analysis differentiates among organisms producing different substances, either directly as molecular product, or indirectly through the use of multivariate statistical data reduction techniques. This method has many advantages over traditional microbial screening methods, which require sample preparation and time for assay development.

The invention provides a method for distinguishing cancer cells through surface enhanced Raman spectroscopy. According to the method, after SERS active nanometer materials and various cells incubated under an ice-bath condition are rapidly preprocessed through ultrasonic, the SERS active nanometer materials are rapidly guided into the various cells through an electroporation method, the surface enhanced Raman spectroscopy of cancer cells is obtained by means of the detection of a Raman spectrometer, a surface enhanced Raman spectroscopy database of the various cells is built, clustering analysis is carried out through multivariate statistics analysis, a distinguishing splattering distribution graph corresponding to the surface enhanced Raman spectroscopy of normal cells and the cancer cells is obtained, and distinguishing of the cancer cells is achieved according to the distinguishing splattering distribution graph. The method for distinguishing the cancer cells through the surface enhanced Raman spectroscopy has the advantages of being rapid and easy to operate, good in universality, low in cost, free of flickering cell SERS spectrums and the like, can achieve large-scale cancer cell detection, and is suitable for being widely used in the technical field of medicine screening, disease diagnosis and the like.

The invention discloses a method of increasing the yield of spinosad by improving the fermentation condition of saccharopolyspora spinosa based on metabonomics. The method comprises the following steps: (1) determining intracellular metabolite of the saccharopolyspora spinosa; (2) analyzing the capability of the saccharopolyspora spinosa of producing the spinosad; (3) carrying out PLS analysis; and (4) carrying out process analysis. According to the method disclosed by the invention, the metabolite change law in the process of producing the spinosad by virtue of fermentation of the saccharopolyspora spinosa is revealed by utilizing the metabonomics means and combining multivariate statistics, and a metabolic change mechanism related to production of the spinosad is further explained; by analyzing the spinosad production capability and the metabolic level of the saccharopolyspora spinosa in different growth environments, the metabolin related to the production of the spinosad is found, thus providing a direction for optimizing the culture process of the saccharopolyspora spinosa and increasing the fermentation yield of the spinosad. The invention also provides new thinking and method for study in other culture technological processes of producing macrolides microorganisms.

The invention discloses a method for screening radiation-damaged biomarkers at the early stage of a rat and an application. The method comprises the following steps of firstly adopting the rat as an animal model, utilizing radiation metabonomics method, adopting a gas chromatography-mass spectrum (GC-MS) combined technology to research the influence of acute ionizing radiation on the change of metabolites of the rat plasma; then utilizing a genetic algorithm (GA) and univariate statistics to analyze and find out potential radiation-damaged biomarkers; finally using the biomarkers for early-stage injury classification of radiation damage by combination with a multivariate statistic analysis method--non-linear iterative partial least square method (KPLS). The invention has the advantages that by adoption of the technical scheme, injury classification can be carried out on a great amount of radiation-exposing crowds in ionizing radiation accidents in a short time so as to have an important significance in early-stage diagnosis and prevention of dysfunction or death of organs; the biomarkers are used for studying the classification of radiation-damaged injuries, so that the limitation of single index for estimation of radiation-damaged dosage is overcome.

The invention relates to the field of artificial intelligence and big data, in particular to a PM2.5 concentration prediction method based on multivariate statistical analysis and LSTM fusion, which comprises the following steps: analyzing the advantages and disadvantages of the two on the theoretical level, and constructing a fusion algorithm based on the two on the basis; acquiring meteorological data and pollutant data from a provincial control point and a national control point; Sampling data in recent half a year, and analyzing the correlation between each factor and the PM2.5 concentration by using a Pearson correlation coefficient; Dividing all the data into three parts, namely training data, test data and prediction data, training a model by using the training data, and setting related model parameters; Inputting test data into the model; According to the method, by analyzing the time and space characteristics of PM2.5, the data is subjected to dimensionality reduction, the deep data characteristics of PM2.5 are mined through the deep learning technology, the data operation speed is greatly increased, prediction work can be carried out in real time by combining with improvement of precision, and the problem of data lagging is solved.

The invention discloses a method of screening a marker of acute renal toxicity caused by aristolochic acid by utilizing cell metabolic profiling in vitro. The effective dose causing renal cell damage is determined by adopting a 3-(4,5-dimethylthiahiazo-2)-2,5-diphenytetrazoliumromide (MTT) staining manner. Renal cell metabolites are analyzed by adopting an ultra-high performance liquid chromatography-mass chromatography technique to obtain metabolic profiles. Metabolic profile data of normal groups and medicine damage groups are analyzed by adopting a multivariate statistic method. A potential renal toxicity marker is screened by combining a change rule of metabolites along with medicine acting time. The screened marker is good in predicting performance. The average accuracy of outer verification is 90% or above. The method comprehensively represents variation situations of metabolites of normal renal cells under medicine damages, and can provide technique supports for early diagnosis of renal toxicity caused by medicines or medicine safety evaluation.

The invention discloses a fractal-wavelet self-adaptive image denoising method based on a multivariate statistic model. The method includes: step one, subjecting a noisy image to homomorphic transform through which an original image IB containing multiplicative noise is transformed into an image IB' only containing additive noise; step two, performing fractal-wavelet transform on a noisy signal f (k), selecting a wavelet basis and a wavelet decomposition layer j to obtain corresponding wavelet coefficients; step three, selecting an MGGD multivariate statistic model for self-adaptive solution of a parameter alpha and a parameter beta, and obtaining the most suitable parameter value alpha and beta after analysis for the distribution condition of the wavelet coefficient of a natural image; step four, for the wavelet coefficients obtained through decomposition, performing noise-free predictive coding on the noisy image by using a fractal-wavelet coding method; and step five, performing wavelet reconstruction by using the wavelet coefficients to obtain estimation signals which are image signals after denoising. Compared with other algorithms, the method has better denoising effect and high edge preserving capacity, and is particularly suitable for eliminating Gaussian-impulse mixed noise.

The invention discloses an electrolytic aluminum whole-process monitoring and fault diagnosis system based on a multivariate statistic method. The system comprises three layers, the bottom layer is used for monitoring key technological parameters, the middle layer is used for monitoring the device running in the electrolytic aluminum process, and the top layer is used for comprehensively monitoring the whole process. The system monitors the bottom layer, the middle layer and the top layer respectively, analyzes measured results comprehensively, and thus guarantees the monitoring pertinence and the accuracy of the monitored results. According to the invention, the separation and integrated monitoring is realized by adopting the concept of monitoring in layers, and further the monitoring pertinence and the accuracy of the monitored results are ensured; forecast before fault, alarming in fault and fault diagnosis and retrospect functions can be realized; the modular design concept is adopted for the off-line modeling and on-line monitoring, and each module performs the corresponding function; not only can the safety production and product quality be ensured, but also the consumption of raw materials and electricity energy can be reduced.

A monitoring test for recurrent breast cancer with a high degree of sensitivity and specificity is provided that detects the presence of a panel of multiplicity of biomarkers that were identified using metabolite profiling methods. The test is capable of detecting breast cancer recurrence about a years earlier than current available monitoring diagnostic tests. The panel of biomarkers is identified using a combination of nuclear magnetic resonance (NMR) and two dimensional gas chromatography-mass spectrometry (GC×GC-MS) to produce the metabolite profiles of serum samples. The NMR and GC×GC-MS data are analyzed by multivariate statistical methods to compare identified metabolite signals between samples from patients with recurrence of breast cancer and those from patients having no evidence of disease.

The present invention discloses an intelligent power distribution network situation perception method Based on multivariate spatio-temporal information modeling. For the first time, a multivariate spatio-temporal information model is used in intelligent power distribution network situation perception which is also a development for the potential connection and inherent characteristics of a power grid so that the state information and measurement information of a power grid can be tapped deeply and effectively. At the same time, the node voltage amplitude with higher precision obtained from the modeling and the node voltage phase angle are taken as virtual measurement information which increases the measurement redundancy of a power distribution network, improves the convergence rate and the convergence precision of the state estimation of the power distribution network, shortens the state estimation time and provides the possibility for the on-line estimation of the intelligent power distribution network. Finally, the security and stability analysis of the real-time and future states of the power distribution network is carried out. The potential risks of the power distribution network are predicted in the future, which provides a reference for system scheduling and system decision-making.