435 results about "Model sample" patented technology

A method for uncertainty estimation for nonlinear inverse problems includes obtaining an inverse model of spatial distribution of a physical property of subsurface formations. A set of possible models of spatial distribution is obtained based on the measurements. A set of model parameters is obtained. The number of model parameters is reduced by covariance free compression transform. Upper and lower limits of a value of the physical property are mapped to orthogonalspace. A model polytope including a geometric region of feasible models is defined. At least one of random and geometric sampling of the model polytope is performed in a reduced-dimensional space to generate an equi-feasible ensemble of models. The reduced-dimensional space includes an approximated hypercube. Probable model samples are evaluated based on data misfits from among an equi-feasible model ensemble determined by forward numerical simulation. Final uncertainties are determined from the equivalent model ensemble and the final uncertainties are displayed in at least one map.

Droplets or plugs within multiphase microfluidic systems have rapidly gained interest as a way to manipulate samples and chemical reactions on the femtoliter to microliter scale. Chemical analysis of the plugs remains a challenge. It has been discovered that nanoliter plugs of sample separated by air or oil can be analyzed by electrospray ionization mass spectrometry when pumped directly into a fused silica nanospray emitter nozzle. Using leu-enkephalin in methanol and 1% acetic acid in water (50:50 v:v) as a model sample, we found carry-over between plugs was <0.1% and relative standard deviation of signal for a series of plugs was 3%. Detection limits were 1 nM. Sample analysis rates of 0.8 Hz were achieved by pumping 13 nL samples separated by 3 mm long air gaps in a 75 μm inner diameter tube. Analysis rates were limited by the scan time of the ion trap mass spectrometer. The system provides a robust, rapid, and information-rich method for chemical analysis of sample in segmented flow systems.

The invention discloses a method of assisting cigarette formula by adopting an SIMCA (Soft Independent Modeling of Class Analogy) based on Near-infrared spectral information, comprising the following steps of: (1) modeling sample preparation; (2) spectrum scanning; (3) spectrum pretreatment; (4) principal component analysis; (5) tobacco raw material database establishment; (6) substitution rule setting; and (7) formula assisting: during the tobacco raw material substitution, scanning the to-be-detected sample according to the steps 1-4 by taking a to-be-substituted tobacco sample as a target, obtaining the near-infrared spectral data after treating, setting the substitution rule on the basis of the information of the substituted sample according to the step 6, comparing the near-infrared data of the substituted sample with a mathematical model in a memory to obtain replaceable tobacco samples, sequencing the samples according to mahalanobis distance, wherein the sample with smaller mahalanobis distance is more similar, and finally carrying out sensory taste. The method disclosed by the invention is reliable and capable of reducing the range of looking for substitute samples for formula personnel, greatly reducing the workload and enhancing the pertinence of tobacco formula work.

The invention discloses a jewelry online design system and an online design method thereof. The jewelry online design system and the online design method thereof are based on digital and visual data processing technologies, and can allow a user to design jewelry styles independently and adding design element symbols flexibly online in a remote manner. The jewelry online design system can view a three-dimensional virtual effect picture of the user wearing the self-customized and designed jewelry in real time, and can carry out three-dimensional printing on the jewelry which is independently designed by the user in a remote manner, so as to output a model sample or commodity entity of the designed jewelry; the jewelry online design system greatly improves design efficiency of the user, is suitable for all types of users to participate since the user does not need to have professional design or drawing knowledge, enhances interactivity, allows the user to select personalized and differentiated portion design elements online from a massive element database flexibly and independently, completes the combined design of rings and other types of jewelry rapidly, and significantly improves the practicality and convenience of jewelry independently designed by the user by means of the functions such as online design, bidirectional scanning and virtual try-on.

The invention relates to the technical field of quality detection and provides a rapid determination method for the evaluation of the storage quality of grain, which comprises: S1: grain samples having different storage years are acquired according to the grain samples to be evaluated, thus obtaining a modeling sample set; S2: a near-infrared spectrum of the modeling sample set is obtained by means of acquisition and is pretreated; S3: a fatty acid value of the modeling sample set is obtained by means of measurement; S4: a correction relation for the modeling sample set is established according to the near-infrared spectrum and the fatty acid value, thus obtaining a correction model; and S5: the storage quality of the grain samples to be evaluated is evaluated according to the correction model. On the basis of near-infrared spectrum technique, the rapid determination method is in no need of the sample pretreatment, fast in detection, great in convenience and simplicity and suitable forthe determination of the fatty acid value in kernel of grain crop such as paddy rice, wheat, corn, soybean and the like, thus technically supporting the real-time detection and surveillance of the storage quality of grain.

The invention discloses a data-driven small watershed real-time flood forecast method comprising the following steps: (1) constructing a delay time sequence and a mean time sequence based on rainfall and river flow data; (2) calculating the grey relevance between the time sequences, and selecting the higher-correlation time sequence as modeling sample data; (3) modeling input-output data by use of a machine learning algorithm to get forecast models and a fitting value; (4) calculating the weight of each forecast model based on the information entropy theory; and (5) forecasting the river flow by use of the forecasting models, and performing weighted calculation on single-model forecast results to get a forecasted value of river flow. According to the invention, the complex nonlinear mapping relationship of small watershed flood is simulated by making full use of rainfall and river flow data, the information in the data is mined, and therefore, the accuracy of real-time small watershed flood forecast is improved.

The invention relates to a connection manner identification and detection technology in an electric power system, and particularly relates to a principal component analysis multivariable decision-making tree-based connection manner identification method. The method comprises four steps of data preparation, decision analysis, decision prediction and model explanation and verification. According to the method, a connection manner characteristic model is constructed on the basis of principal component analysis multivariable, and a connection manner characteristic model sample is trained by adopting the thought of the decision-making tree so as to generate a classifier, so that the connection manner identification under unknown condition is realized.

The invention relates to a nerve network clustering method based on iteration. The method comprises steps that 1, parameters of an over-limit learning machine model are initiated; 2, samples having the same number as the required clustering number are randomly selected, each sample represents one clustering, an initial model sample set is formed, and an initial over-limit learning machine model is acquired through training; 3, clustering grouping for the samples is carried out by utilizing the initial over-limit learning machine model, and a clustering result is acquired; 4, for each clustering group, multiple samples are selected according to rules as models of the clustering groups; 5, the model samples of each clustering group acquired in the step 4 are utilized to update the over-limit learning machine model; and 6, the process returns to the step 3 for iteration till the clustering groups are stable or iteration frequency requirements are satisfied, and the clustering groups are acquired and outputted. The method solves a high-dimensional non-linear data space clustering problem and further solves problems of large memory consumption and long operation time.

The invention relates to an online monitoring method for resistance spot welding quality and belongs to the field of resistance spot welding quality monitoring techniques. At first, the welding current and inter-electrode voltage for resistance spot welding are collected in real time in the actual production process of resistance spot welding, a dynamic resistance curve is worked out, according to the welding process application principle in actual production, namely a qualified welding spot is a great probability event, a system carries out modeling sample capacity estimation according to the principle of statistics, cluster analysis is conducted on the dynamic resistance curve, and the qualification probability of a new welding spot is predicted according to the analysis result. Modeling is not needed under a specific welding process condition, an automatic modeling mode has the characteristic of universality, welding production data are directly adopted for modeling, and therefore the online monitoring method is more suitable for actual production and greatly improves online monitoring accuracy.

The invention provides a vehicle number recognition method and device. The method comprises the following steps of respectively carrying out normalization on each of a plurality of pre-obtained vehicle model sample pictures to obtain a plurality of standard sample pictures; obtaining a character convolutional neural network structure according to a vehicle number character training neural network in the standard sample pictures; obtaining an end position convolutional neural network structure according to an end position character training neural network in the standard sample pictures; and respectively utilizing the character convolutional neural network structure and the end position convolutional neural network structure to recognize the pre-collected vehicle model pictures to obtain a vehicle number. The device comprises a normalization module, a character training module, an end position training module and a recognition module. According to the vehicle number recognition method and device, the vehicle number is obtained by utilizing the trained character convolutional neural network structure and the end position convolutional neural network structure, so that the accuracy rate for recognition is improved.

The invention discloses an electronic device, a multi-model sample training method and a computer readable storage medium. The method comprises the following steps of receiving sample data uploaded bya user, and determining the data attributes of the sample data, wherein the data attributes comprise types and numbers; according to the mapping relation between the preset data attributes of a machine learning model and the sample data, determining a machine learning model corresponding to the data attributes of the sample data; training the sample data respectively by the determined machine learning model; analyzing the training result of each machine learning model after being well trained; and displaying the training result conforming to a preset condition on a display interface. According to the technical scheme, the situation that the training result is poor due to the inadequate experience of a user and the wrong selection of a model can be avoided. The problem that the working progress is influenced due to the fact that the user spent too much time to repeat training the model is solved.

A method for modeling samples includes the use of control points to define lines profiles and other geometric shapes. Each control point used within a model influences a shape within the model. Typically, the control points are used in a connect-the-dots fashion where a set of dots defines the outline or profile of a shape. The layers within the sample are typically modeled independently of the shape defined using the control points. The overall result is to minimize the number of parameters used to model shapes while maintaining the accuracy of the resulting scatterometry models.

The invention relates to an automatic recognizing method and device for nitrogen phosphorus and potassium stress of protected tomatoes. The automatic recognizing device consists of a CCD (charge coupled device), a spectrograph, an optical system, a light source, a displacement platform, a controller, a PC (personal computer), a display and a light box. The method comprises the following steps of: firstly, performing system experiment calibration on the automatic recognizing device by a calibrating sample; then sampling a model sample during a whole growth period at equal time intervals, and collecting the reflection intensity, the polarization distribution and hyper spectrum image information of visible light-near infrared of inverted-7-shaped leaves of the tomatoes; analyzing and acquiring a main ingredient image of a nutrition stress sample according to main ingredients by combining with a measured value of chemical nitrogen phosphorus and potassium analysis and a micro structure scanning electron microscope test of the model sample, thus determining the characteristic wavelength of nitrogen phosphorus and potassium; and classifying nutrition stress characteristics into three types including a polarization characteristic, an intensity characteristic and an image characteristic, establishing the characteristic space of the nutrition stress characteristics, and establishing an automatic recognizing and comprehensive evaluation model for nitrogen phosphorus and potassium of the tomatoes according to a D-S (Dempster/Shafer) criterion theory, thus automatically recognizing and diagnosing the nitrogen phosphorus and potassium stress types of the protected tomatoes.

The invention provides a method for conducting statistics on elevator visitor flow based on intelligent visual perception. The method comprises the steps of S1. establishing a head-and-shoulder model sample database; S2. conducting feature extraction and model training, wherein samples in S1 are subjected to principle component analysis (PCA) feature extraction, and a support vector machine (SVM) trainer is used for training model generating; S3. detecting targets, wherein matching calculation is conducted on images collected in real time according to human body head-and-shoulder model data obtained in S2, and targets in current images are obtained through detection; S4. tracking targets, wherein targets detected in S3 are tracked; and S5. conducing statistics on the visitor flow, wherein corresponding counters are operated according to incoming and outgoing conditions of targets when targets tracked in S4 go cross a crossing line. The method has the advantages that real-time intelligent analyses are conducted on images collected in real time, so that elevator visitor flow data can be obtained, accurate evidences are provided for establishing of effective and energy-saving elevator dispatching strategies, the problem that prediction models of the elevator visitor flow in traditional methods are complex and difficult to establish is solved, and the imprecise prediction caused by special events are prevented.

The invention relates to the technical field of human-computer interaction, and particularly relates to an information processing method and apparatus in the human-computer interaction. The information processing method includes providing a model sample database including a sample standard question and a sample extension question corresponding to each sample standard question; providing a knowledge base including a knowledge base standard question and a knowledge base extension question and answer corresponding to each knowledge base standard question and used for providing answers for users' questions; determining whether there is a sample extension question that matches the user's question in the human-computer interaction log in the model sample database; if yes, determining whether the corresponding standard question of the user question in the human-computer interaction log is the same as the corresponding sample standard question of the matched sample extension question; if no, optimizing the knowledge base. The invention further provides an information processing apparatus and system corresponding to the information processing method.

The invention discloses an intelligent early warning method for dam safety monitoring data. The method comprises the steps of early warning model establishment, threshold value setting and mutual feedback type early warning. Gross error identification and gross error processing are carried out, model sample data quality is improved, according to the monitoring items, independent variable relevance, historical monitoring data quantity and historical monitoring data distribution, different early warning models and indexes are established, including a stepwise regression model, a correlation vector machine model and a gray system model; the established models can reflect the relationship between the independent variable and the dependent variable more truly and are wide in application range,according to a measuring instrument, measuring point attributes, a threshold value, an early warning model and indexes, real-time early warning is carried out on monitoring data, monitoring instrumentabnormity early warning is sent to monitoring personnel, or dam safety early warning is sent to dam safety management personnel, experts with professional knowledge and rich experience are not needed, the workload is small, the early warning speed is high, and the early warning result is more accurate and reliable.

The invention discloses a motion detection optimization method based on a ViBe background modeling algorithm. A novel method is used for obtaining a group of background models, the novel background models have time information and space information simultaneously, the false detection rate can be lowered, and the error rate of background updating can be reduced; an ideal background model is composed of true background points, and therefore it is hoped to obtain a background model with credibility as high as possible in the background updating process. According to the method, samples of the background models are divided into a group with high credibility and a group with low credibility through a simple method, the samples with the high credibility are reserved, and the background models with the low credibility are updated randomly; by the adoption of the updating method, the credibility of the reserved background model samples can be higher and higher; in terms of postprocessing, filtering processing is added in addition to morphological operation, and abnormal points in a foreground image are further removed.

The invention discloses a weighted least square support vector machine-based soft measurement method for magnetic flux linkage of a bearingless permanent magnet synchronous motor. The method comprises the following steps of: selecting a rotor position angle, torque winding current, levitation force winding current and rotor eccentric displacement of the bearingless permanent magnet synchronous motor as four input variables of a bearingless permanent magnet synchronous motor magnetic flux linkage soft measurement model, wherein the magnetic flux linkage y is taken as an input variable; acquiring representative input variable sample data and output variable sample data, performing normalization preprocessing on both the output variable and the input variables, and forming a modeling sample set for corresponding normalized values; performing residual analysis on the modeling sample set to acquire each sample weight; training the modeling sample set and establishing a bearingless permanent magnet synchronous motor magnetic flux linkage correlation model by using a weighted least square support vector machine; and finally abnormalizing an sy value to acquire the magnetic flux linkage y. Thus, the magnetic flux linkage value of the bearingless permanent magnet synchronous motor is predicted and controlled on line in real time.

The invention discloses a method for building a near-infrared qualitative analysis model. The method includes the steps of (1) collecting near-infrared spectroscopic data of a sample, and determining modeling sample data; (2) preprocessing the modeling sample data; (3) conducting partial least square feature extraction on the modeling sample data; (4) conducting orthogonal linear discriminant analysis feature extraction on the modeling sample data; (5) building the qualitative analysis model with a support vector machine method. The method is accurate, efficient, rapid and free of professional operation.

The invention relates to a method for determining explosion limit of multicomponent mixed gas. The conventional physical and chemical parameters and the mixture ratio of pure gas are adopted to build corresponding explosion limit prediction models aiming at different classes of multicomponent mixed gas, and the built models are utilized to predict the explosion limit of unknown mixed gas. The method comprises the following steps of: 1, collection of modeling samples of the multicomponent mixed gas and the explosion limit data of the multicomponent mixed gas; 2, classified processing of the modeling samples of the multicomponent mixed gas; 3, determination of the physical and chemical parameters and collection of the data; 4, building of the prediction models; 5, confirmation and correction of the models; 6, and application of the prediction models. The explosion limit is an important indicator for measuring the explosion risk of the mixed gas, and a predicting method for researching the explosion limit of the multicomponent mixed gas has importance for the prediction and monitor of the gas explosion. The method is sample and has high prediction accuracy, and the invention provides a simple, fast, accurate and reliable predicting method for predicting the explosion limit of the multicomponent mixed gas.

The invention discloses a rapid fruit texture non-destructive detection method and a non-destructive detection device. The detection method comprises the following steps: acquiring a modeling sample set; determining the quality; determining the fruit shape coefficient; acquiring vibration response data; extracting vibration characteristic parameters; performing dimension reduction on the vibration characteristic parameters; determining a texture index value; and building a texture prediction model and measuring. The detection device comprises a vibration control system and a vibration signal acquisition system. According to the detection method and the detection device, an impact vibration mode is adopted to greatly shorten the fruit texture detection time by use of the laser Doppler vibration measurement technology, multi-vibration characteristic parameters combined with the fruit quality and the fruit shape are used as input parameters of the prediction model, and the prediction accuracy and the adaptability of the model can be improved.

The invention provides a detection method of gasoline properties based on weighted absorbance and similar samples, the detection method comprises the steps of calculating the correlation coefficient R between a sample to be detected and a known sample on basis of the gasoline properties and the absorbance of the known sample after the near infrared spectroscopy of the sample to be detected and the known sample is conventionally pretreated; taking the correlation coefficient R as a weight, the weighted sample absorbance is obtained and the weighted absorbance matrix is established; calculating the score matrix of the weighted absorbance matrix by adopting a principal component analytical method; selecting the first principal component and the second principal component in the score matrix to obtain a new score matrix; calculating the mahalanobis distance according the new score matrix, selecting the similar samples which are nearest to the mahalanobis distance of the sample to be detected in the known spectral database as modeling samples, and performing prediction on the sample to be detected. According to the detection method provided by the invention, the effect of gasoline properties on the spectral distance is fully considered, the phenomenon that the classification is not accurate can be effectively avoided, so that the prediction accuracy of a model is finally improved. An important guarantee can be provided for accurately measuring the gasoline properties and timely adjusting the operating parameters.

The invention discloses a method for modeling a relationship between chemical wood pulp cellulose characteristics and paper sheet tensile strength. The method is characterized in that wood pulp cellulose, which is prepared according to a chemical method, is taken as a model sample; the wood pulp cellulose comprises softwood pulp and hardwood pulp; 12 types of fibers with different morphological parameters are obtained by means of performing defibering and PFI defibrination on different types of wood-pulp boards; for a paper sheet in the model, a tensile strength is assumed as T, an Elongation is assumed as s, and a fiber crimp index is assumed as Cw, then a paper sheet tensile strength model expression is T=m*Cw*S, wherein a, m and n are constant numbers. According to the method for modeling the relationship between the chemical wood pulp cellulose characteristics and the paper sheet tensile strength, all parameters in the paper sheet tensile strength model are fiber characteristic parameters, and can be obtained through conventional analysis and test methods.