269 results about "Multiple linear regression model" patented technology

The invention discloses a fish tenderness hyperspectral detection method based on a characteristic wave band. The detection method comprises the following steps of: firstly, detecting fish tenderness according to the conventional detection method; then, with fish as a sample set, utilizing a near-infrared hyperspectral imager to carry out nondestructive detection on the fish, establishing a multi-element linear regression model of the tenderness and a characteristic wavelength, detecting a fish sample, and calculating the tenderness of fish sample according to the multi-element linear regression model of the tenderness and the characteristic wavelength. The fish tenderness hyperspectral detection method provided by the invention has the advantages that the fish tenderness is rapidly detected in an optimized manner through the characteristic wave band, the time of hyperspectral collection, detection and data analysis is greatly shortened, the detection efficiency is improved, and the fish tenderness hyperspectral detection method really provides theoretical support and research basis for on-line, rapid and nondestructive detection of food quality.

The present invention discloses a zone area reasonable line loss prediction method based on a data mining technology. The beneficial effects of the present invention are that: the method of the present invention aims to discover the potential correlation between the zone area characteristic data and the zone area line loss by using the actual sampling data of the zone area line loss and utilizing a data mining means to process the mass data, is reliable in data source by being compared with a conventional theoretical line loss calculation method, and can reflect the on-site actual line loss status better; the algorithm flow is simple and practical, and the efficiency is higher; a multiple linear regression model is compact in structure, can reflect the influence degrees of the characteristic parameters to the zone area line loss visually, and is easy to analyze the status of a zone area of unreasonable line loss; the method is easy for software realization, and can be integrated in a line loss management system more conveniently.

The invention belongs to the technical field of typhoon intensity prediction, and specifically relates to a typhoon intensity prediction method and a system. The typhoon intensity prediction method includes: step a: statistics and analysis of historical typhoon data; step b: selecting an environmental prediction factor related to typhoon intensity change, and calculating and extracting parameter data related to the environmental prediction factor; and step c: building a multiple linear regression model, taking the parameter data related to the environmental prediction factor of the historical typhoon data into the multiple linear regression model for regression model training, obtaining a typhoon intensity prediction equation according to a training result, and calculating the typhoon intensity according to the typhoon intensity prediction equation. According to the typhoon intensity prediction method and the system, not merely basin typhoon can be predicted, all types of typhoon can be predicted, the prediction range is wider, the selection of the environmental prediction factor is more comprehensive, and the accuracy of typhoon intensity prediction is improved.

The invention discloses a river and lake water quality monitoring method based on high-resolution satellite images. The river and lake water quality monitoring method comprises satellite image radiometric calibration, atmospheric correction, RPC orthographic correction, image splicing, automatic water body extraction, water quality quantitative inversion modeling, water quality quantitative inversion model precision verification and water quality quantitative inversion model application. The invention relates to the technical field of inland water environment remote sensing science, in particular to a river and lake water quality monitoring method based on a high-resolution satellite image, which has the advantages of wide monitoring range, high speed, low cost and convenience in long-termdynamic monitoring by utilizing a remote sensing technology. A multiple linear regression model between different water quality parameter concentrations and image waveband reflectivity is establishedby combining sampling data of a water quality monitoring station and representation of various substances in a water body on a remote sensing image, and a relative error and an absolute error of themodel are calculated according to an inversion result, so as to promote the model, and water quality evaluation of the whole river and even a water area with a larger range is achieved.

The invention provides a construction method of a software defect evaluation model on the basis of a complex network, and can predict potential defects in a software system. The construction method comprises the following steps: 1) taking classes in software as nodes and a relationship between the nodes as edges to construct the directed network model of the software; 2) according to the directed network model obtained in 1), carrying out the characteristic measurement calculation of the complex network; 3) scanning a software source code to carry out structured program measurement, and obtaining the measurement values of cyclomatic complexity and function depth; 4) scanning the software source code to carry out object-oriented software measurement; 5) utilizing a FindBugs static analysis tool and a software defect report on an open-source tool official website to search and analyze software defect information; and 6) constructing a defect evaluation formula: using the calculation data of a corresponding effective measurement index for different classes of software, establishing a corresponding multiple linear regression model, and obtaining the software defect evaluation model.

The invention discloses a spectral morphological characteristic-based hyperspectral water quality parameter quantitative inversion method, relating to the field of water quality remote sensing. The method comprises the steps of comparing and analyzing ground measured spectral data and hyperspectral image data, extracting a spectral curve morphological characteristic, selecting a ground measured spectral morphological characteristic significantly correlated to water quality parameters and building a ground measured spectral data-based inversion model, building a hyperspectral inversion model ofeach water quality parameter with a spectral morphological characteristic selected by the inversion model built based on the ground measured spectral as an independent variable, and applying the hyperspectral inversion model to a hyperspectral image to acquire a water quality parameter inversion result of a working area. Through adoption of the method, multivariate regression models of common chemical water quality parameters such as pH and hardness can be built, multiple types of water quality parameter information can be acquired from point to surface rapidly and accurately, and a new technological method is provided for regional water environment dynamic monitoring.

The invention relates to fine granularity area estimation of air pollutant, in particular to a fine granularity air pollutant concentration area estimation method based on spatial characteristics. The method comprises a preprocessing stage, a predicting stage and an estimating stage. In the preprocessing stage, mesh generation is carried out to obtain a train sample. In the predicting stage, a distance measure learning method is used for learning the mahalanobis distance function to find k areas in the shortest distance, and multiple linear regression training is carried out on characteristic vectors formed by the k areas. In the estimating stage, a multiple linear regression model obtained through training is used for estimating the air pollutant concentration in the areas. The method has the advantages that the air pollutant concentration of some small areas without air quality monitoring stations can be estimated, adjacent air quality monitoring stations are fully utilized, the k adjacent areas are introduced, the fine granularity change of the air pollutant concentration in space can be better captured, and the pollutant concentration of the current area can be estimated more accurately.

The invention discloses a power distribution network operation efficiency evaluation method and system based on big data mining, and relates to the field of the power distribution network monitoring and control. The method comprises the following steps: collecting a power distribution network operation efficiency evaluation parameter, and establishing a power distribution network operation efficiency evaluation parameter database; constructing at least a multiple linear regression model, an exponential smoothing model, a first-order / second-order self-adaptive combination model and a neural network parameter optimization model; and evaluating and / or predicting the power distribution network operation efficiency by using the multiple linear regression model, the exponential smoothing model,the first-order / second-order self-adaptive combination model and the neural network parameter optimization model. By mining the relativity big data of the power distribution network, the distributionnetwork equipment and system operation efficiency, the coordination and the equipment equilibrium degree, the public power grid capital, and the user dedicated capital efficiency are analyzed according to different power supply regions and function zone types, and the influence factors with low efficiency are mined, thereby forming an efficient tool for distribution network efficiency monitoring,and a certain tendency pre-judgment and estimation function is provided.

The invention discloses a solar irradiation predicting method based on multiple linear regression. The method comprises the first step of calculating the theoretical value of the solar irradiation value, the second step of calculating the attenuation rate of the solar irradiation, the third step of establishing a month-by-month regression equation of the cloudage, the cloud picture brightness temperature, visible light emissivity and the radiation attenuation rate based on a multiple linear regression model, and the fourth step of calculating the actual solar irradiation intensity predicted value according to the month-by-month regression equation established in the third step and the theoretical value of the solar irradiation value. The month-by-month regression equation of the cloudage, the cloud picture brightness temperature, the visible light emissivity and the radiation attenuation rate is established based on the multiple linear regression model, the characteristic quantities of a cloud picture are predicted according to the linear regression model, and the purpose of accurately predicting the solar irradiation is achieved.

The invention provides a low-cost calibration method for PM2.5 monitoring nodes. The method includes the following steps that the nodes are deployed near an air quality inspection station, and training samples consistent in time and space are obtained; models are built to show the relationship between a number read at each node and a PM 2.5 true value; training data is preprocessed, wherein parts of characteristics are standardized and a training sample set and a testing sample set are determined by means of a set-aside method; for a linear invariant model, a three-layered back-propagation neural network is adopted to train a multiple linear regression module on the training sample set, and verification of accuracy of the models is completed on the testing sample set; for a linear variable model, in the time interval, the training samples are fitted simply by means of the least square method to obtain a linear parameter, in different time periods, linear parameters, average values of node readings and average values of the sensitive characteristics data serve as new training samples, post-pruning-strategy-based CART regression tree training is adopted on the new training samples, and verification of reliability of the models is completed on the testing sample set; an off-line model which is verified to be accurate is written to a node program.

The invention provides a method and a device for determining the content of adsorbed gas in a clay shale reservoir and belongs to the technical field of prediction of gas content of the clay shale reservoir. The method comprises steps as follows: influence factors of the adsorbed gas in the clay shale reservoir are obtained; the actually measured adsorbed gas content is acquired through a preset adsorption model according to physical property parameters and isothermal adsorption experimental data of a to-be-measured clay shale reservoir; a scatter diagram indicating that the actually measured adsorbed gas content changes along with the adsorbed gas influence factors is determined according to the adsorbed gas influence factors and the actually measured adsorbed gas content; the scatter diagram is subjected to fitting, a correlation curve and the fitting degree of the adsorbed gas content changing along with the predetermined influence factors are acquired, and the influence factors with the fitting degree larger than a preset value are determined as main controlling factors of the adsorbed gas content; a multiple linear regression model is established according to the main controlling factors of the adsorbed gas content; an adsorbed gas content model is determined through the multiple linear regression model according to the adsorbed gas influence factors and the actually measured adsorbed gas content. The method and the device have the characteristic of higher accuracy of a quantitative prediction result.

The invention discloses a prediction method for establishing a multiple linear regression model on the basis of a grey correlation analysis method so as to predict a coal calorific value. The method carries out correlation analysis on the coal calorific value and five indexes including moisture, ash content, volatile components, gelatinous layer maximum thickness and an oxygen and carbon atomic ratio to find a main impact factor associated with the coal calorific value and establish the multiple linear regression model so as to predict the coal calorific value. The method adopts a correlation analysis method in a grey system theory to analyze five factors which affect the coal calorific value, a main factor which affects the coal calorific value is picked up from the five factors, and the multiple linear regression model between the coal calorific value and the main impact factor is established. The prediction method of the coal calorific value is simple and feasible and is high in prediction precision, and a relative prediction error does not exceed + / -8%.

The invention relates to a low voltage grid-connected detection device and method of a distributed new energy power generation system. The low voltage grid-connected detection device comprises a signal gathering module, a signal modulation module, a master control module, a wireless communication module and a grid-connected control module. The signal gathering module comprises six alternating voltage transformers, three alternating current transformers and a direct voltage transformer. The signal modulation module comprises a three-phase filter circuit, a three-phase voltage modulation circuit, a three-phase current modulation circuit and a zero cross detection circuit. The master control module comprises a comparator and a DSP processor, wherein the comparator comprises a voltage comparator, a phase angle comparator and a frequency comparator. A new energy power generation system grid-connected rank function is built by a multiple linear regression model, the new energy power generation system grid-connected rank function synthesizes eight parameters relative to the quality of electric energy, and therefore obtained rating results are more accurate and comprehensive. According to the low voltage grid-connected detection device and method, detection results are uploaded to a power grid dispatching center by the wireless communication module in time, the plug-and-play characteristic is achieved, and the grid-connected rate of the distributed new energy power generation system is improved.

The invention discloses a power short-term load prediction method and device. The method comprises the steps of dividing day types of historical data based on a daily load curve; Respectively establishing a plurality of multivariate linear regression prediction models aiming at different day types by taking the obtained historical daily load information contained in each day type and the selectedcharacteristic data as input and taking the predicted daily load value as output; And based on the TensorFlow deep learning model, carrying out training, parameter tuning and verification on the plurality of established multivariate linear regression prediction models to obtain short-term load prediction models for different day types. According to the invention, day types can be automatically divided according to information such as holidays and daily load curves; A multivariate linear regression model is adopted, and influences of holidays and weather changes on loads are comprehensively considered; And automatically training and optimizing according to the daily type under a deep learning framework to obtain three short-term load prediction models, and calculating to obtain a relativelyaccurate load prediction value.

The invention relates to an undirected network edge-connectivity weight prediction method based on node similarity. The undirected network edge-connectivity weight prediction method comprises following steps of 1) establishing an undirected network diagram in dependence on an existing data set including nodes and the edge-connectivity weight; 2) respectively calculating following three kinds of similarity indexes including the local similarity index, the global similarity index and the semi-local similarity index in the network diagram in step 1); and 3) predicting the edge-connectivity weight in the test set in dependence on the three kinds of similarity indexes calculated in the step 2) and by means of a multivariate linear regression model, and then examining the model mean performances including the Pearson coefficient and the root-mean-square value by means of a tenfold crossing verification method. According to the invention, by means of the node similarity, the missing edge-connectivity weight is predicted through the multivariate linear regression model, the model is simple, and the prediction result is good.

Methods and systems for calculating body fluid metrics are provided. In accordance with an exemplary embodiment of the present technique, there is provided a method for calculating body fluid metrics by acquiring an absorbance spectrum of a subject's tissue over a range of near-infrared light, performing a multi-linear regression of the absorbance spectrum of the subject's tissue in relation to absorbance spectra of tissue constituents, and calculating body fluid metrics based on the results of the multi-linear regression. A system is provided having a sensor for emitting the light into the subject's tissue and detecting reflected, scattered, or transmitted light, a spectrometer for processing the detected light and generating the absorbance spectrum of the subject's tissue, memory for storing absorbance spectra of the tissue constituents and a multi-linear regression model, and a processor for performing the multi-linear regression and calculating the body fluid metrics.

The invention belongs to the technical field of food, and more specifically relates to a color / fluorescence double signal visible rapid nitrite detection method, and applications thereof. The color / fluorescence double signal visible rapid nitrite detection method comprises following steps: firstly, a carbon dot-neutral red composite system is prepared, wherein carbon dot synthesis and constructionof the carbon dot-neutral red composite system are carried out; a common filter paper is immersed in the carbon dot-neutral red composite system so as to obtain a carbon dot-neutral red composite system detection test paper; a sodium nitrite standard solution is prepared; a color standard colourimetric card and a fluorescence standard colourimetric card of nitrites are prepared; a software is adopted to extract the RGB values of each test paper, a matrix is constructed, and a multivariate linear regression model is constructed; and the content of nitrites in food to be detected is detected. The carbon dot-neutral red composite system possess color / fluorescence double response capacity on nitrites; detection operation steps are simple; the visible results are clear, visual, and accurate; the detection limit is lower; and it is promising for the color / fluorescence double signal visible rapid nitrite detection method to be used in large-scale food nitrite content detection.

The invention provides a streaming data self-adaption persistence method and system based on mixed storage. The method includes the steps that state feature information of a streaming data processing system is collected in real time; a multiple linear regression model based on machine learning is established, and model parameters are estimated according to the collected state feature information; the optimal persistent window size of the streaming data processing system under the current state is calculated and obtained according to the state feature information of the current streaming data processing system and the established regression model; the streaming data processing system changes the current persistent window according to the obtained persistent window size, and the middle state or the calculation result in the streaming data processing process is stored in a solid state disk; when data capacity in the solid state disk reaches a certain degree, data in the solid state disk are stored in an ordinary hard disk. By means of the method and system, the persistent window size at the moment can be calculated according to the current and historical state information, accordingly the situation that the streaming data rate is unstable is dynamically adapted, and balance between usability and consistency of the system is guaranteed.

The invention provides a catering friend edge influence prediction method based on multivariate linear regression, and belongs to the field of data mining. Firstly a friend network diagram is established according to the friend relationship; then the topological characteristics in the friend network are extracted; the general characteristics of the user are considered, and non-topological characteristic information in the friend relationship is extracted; and connection of each characteristic and the final user friend relationship is obtained by using a multivariate linear regression model. The method comprises the steps that the characteristics are selected according to the data set, and a prediction model is established by using the linear regression method. According to the catering friend edge influence prediction method based on multivariate linear regression, the topological characteristics in the friend network and the overall non-topological characteristics of the user are integrally considered so that the friend edge influence can be predicted.

The invention discloses a database mode abstract generation method based on label propagation, and belongs to the field of the application of the database technology. The database mode abstract generation method based on label propagation comprises the following basic steps that the importance of each relation table is measured through a mapping method from a database mode to a label graph according to the information of a primary key and a foreign key, attribute information and tuple information in the relation table; the similarity of the relation tables is calculated through a multiple linear regression model, the mapping relationship of names, attribute values and relation tables is selected to serve as the main characteristic of a similarity model; mode information is clustered through a semi-supervised label propagation algorithm, and a mode abstract is generated automatically. By means of the database mode abstract generation method based on label propagation, an effective solution can be provided for automatic generation of a large-scale database mode abstract, and the purpose of helping a user rapidly understand the database mode information is achieved.

The invention discloses a city expansion boundary prediction method based on space syntax; the method comprises the following steps: 1, drafting a segment map and extracting a built up area boundary; 2, setting up a fixed grid unit; 3, calculating space syntax form analysis variables, and assigning values to the grid; 4, calculating distances from each mesh to the built up area boundary; 5, setting up a multivariate linear regression model and examining the model; 6, predicting a city expansion boundary. The method forms the built up area boundary and the space syntax form analysis variables quantification model, and applies the model in the city expansion boundary prediction, thus providing novel thinking for city program decision and city boundary problem research.

The invention discloses a cotton color measuring method. The method comprises the following steps: 1, setting up a cotton color measuring hardware platform; 2, carrying out single correction of a color difference meter; 3, detecting a cotton sample; and 4, utilizing the optical switching multiple linear regression model stored in the color difference meter for calculation to obtain the cotton rating parameter indexes comprising the Rd color space and the +b color space. The method allows the cotton color to be determined through utilizing a universally accepted color measure standard by establishing the multiple linear regression model of the L*, A* and B* information quantification in an CIE three-dimensional color system based on the Rd and +b in an HVI color system, so cotton color measuring apparatuses can be substituted by testers common in the market, thereby miniaturization and portability are realized, and the determination environment is not limited to the lab environment.

The invention discloses a method for detecting the chlorophyll content and the biomass of chlorella based on a spectrum technology. The method comprises the steps as follows: (1) collecting spectral transmittance values of a chlorella reaction liquid sample under the wavelengths of 439 nm, 543 nm, 696 nm and 1065 nm, and measuring the chlorophyll content of the sample; (2) creating a multiple linear regression model by taking the spectral transmittance values as input vectors and the measured chlorophyll content as an output vector, obtaining the biomass of the sample, and creating a correlation between the chlorophyll content and the biomass; and (3) collecting spectral transmittance values of chlorella-to-be-detected reaction liquid under the wavelengths, and substituting the spectral transmittance values into the model to work out the chlorophyll content and the biomass of the chlorella-to-be-detected reaction liquid. According to the method, the chlorophyll content and the biomass of the chlorella can be quickly and accurately detected, the operation steps are greatly reduced, the detection time is shortened, and consequences such as inaccurate measurement results caused by unskilled operation of an operator or subjective factors are also avoided.

The invention relates to an information mining and progress forecasting method based on heterogeneous system integration. Program or task progress evaluation of a plurality of current heterogeneous systems mainly has the following problems: 1, integrated integration and information sharing of the heterogeneous systems cannot be achieved; 2, the information mining depth of the progress of the various systems is not enough; 3, and no complete progress forecasting method is provided. By means of the information mining and progress forecasting method, integration of heterogeneous information systems is achieved, further a data warehouse is built, corresponding program or task progress information can be obtained through data mining, the obtained progress information is substituted into a multiple linear regression model, an accurate mathematical model can be obtained through optimization, the progress of the program or task can be calculated through the model, and the displayed graph can be audio-visual and easy to read through the database base layer technology and C++Builder visual control. The information mining and progress forecasting method introduces the multiple linear regression model into the progress forecasting algorithm of the program or task and improves the accuracy in progress forecasting.

The invention relates to a calibration method for micromagnetic detection of ferromagnetic material structural mechanics performance, and belongs to the technical field of micromagnetic non-destructive detection. The method includes the steps of selecting samples, wherein a sample not detected and a sample determined to be waste are randomly selected from a production line and a library for parts determined to be waste to serve as a checking sample and a calibration sample respectively, micromagnetic measurement and a conventional mechanics performance test method are carried out respectively, a multiple linear regression method is adopted, and a linear combination equation Y=F(X) formed by micromagnetic parameters is given for each mechanical property parameter; checking model prediction accuracy, wherein the micromagnetic parameters of the checking sample are substituted into a multiple linear regression model to obtain an estimation result of the mechanical property parameters, the error between the estimation result and a conventional measurement result is calculated, and if the error is smaller than a permissible error defined in advance, calibration is completed; otherwise, the steps are repeated. Micromagnetic detection is carried out on samples to be detected, wherein the samples are made of the same material through the same technological process, the obtained micromagnetic parameters are substituted into a multiple linear regression equation set, and the mechanical property of the sample to be detected can be obtained.

A method for optimizing the operating parameters of the xylene isomerizing reactor includes such steps as combining the radial basic function RBF network of multi-variable interpolation with PLSR, inserting the default output generated by various sample data in the equation to obtain a regression model, using PLSR method to find out the regression question, choosing the ethylbenzene transform rate, isomerizing rate, C8 arylhydrocarbon output rate, etc as the dependent variables of said model, using the actual data of industrial reactor as the training samples, and performing the calculation to find out the optimal parameters.

The invention discloses a hail damage remote sensing monitoring method for agricultural insurance claims, relates to a damage remote sensing monitoring technology, and concretely relates to an application of the remote sensing technology in hail damage monitoring and an application of the remote sensing technology in agricultural insurance. The problems of information asymmetry, low loss assessment efficiency and many claims disputes in traditional agricultural insurance survey and loss assessment working processes are solved in the invention. The remote sensing monitoring method includes thefollowing steps: 1, obtaining a reflection spectrum; 2, performing spot test on the crop loss rate, and recording the GPS information of a sampling point; 3, constructing a multiple linear regressionmodel; 4, inverting the loss rate of a hail damage land; and 5, performing claim settlement on the land with loss rate in a hail damage range of 30% or more. The remote sensing technology is combinedwith a traditional agricultural insurance survey and loss assessment working mode to create a novel remote sensing technology application mode and a novel survey and loss assessment mode.

The invention discloses a method for evaluating sweet potato quality on the basis of a vibration sound signal. A sweet potato moisture model is as shown in the specification; a sweet potato total sugar model is as shown in the specification; a sweet potato amylose model is as shown in the specification; and a sweet potato amylopectin model is as shown in the specification. The method evaluates thesweet potato quality through a multivariable linear regression model established by a time frequency characteristic value of the vibration sound signal, and a novel method is provided for a food quality evaluation means.

The invention discloses a VoIP voice quality objective evaluating method based on network parameters, and belongs to the voice quality evaluation research field. The method includes the following steps: obtaining network parameters during VoIP conversation; pre-processing the acquired network parameters, pre-processing including obtaining basic main statistics such as mean value, variance, extreme value, mode and median, wherein the network parameters mainly refer to a packet loss ratio, a jitter value and a time-delay value; performing PCA dimensional reduction on the preprocessed data; and finally inputting the data undergoing dimensional reduction to a multivariable linear regression model, and obtaining a voice quality objective evaluation score through calculation. In this way, evaluation of the VoIP voice quality based on the network parameters is conducted. The invention brings forward the new VoIP voice quality objective evaluating method, which can accurately evaluate the voice quality only by monitoring network parameter values with no need of obtaining voice signals.

The invention discloses a multiple linear regression model-based belt weigher main error factor analysis method. The method includes the following steps that: a belt weigher experiment platform is set up, main error factors such as tensions, temperatures and equivalent flow rates are changed according to actual situations, the tensions, temperatures, equivalent flow rates, calibration values and hopper weighers of various test points are recorded, and the relative errors of the calibration values and hopper weighers are calculated; the correlation coefficients of the main error factors of a belt weigher are calculated, and the correlation of the main error factors is judged; the relations of the factors are determined according to the correlation, and a multiple linear model is set and solved; residual sum of squares, determination coefficients and MS variance of residuals are adopted as check indexes, and the regression effect of the established model is evaluated; and the test values of the test points are predicted through using a fitting model, and the test values are compared with practical values, so that prediction errors can be calculated, and the accuracy of a fitting result is determined. With the method of the invention adopted, a theoretical basis is provided for quantitative analysis on the degree of influence of the main error factors on the accuracy of the belt weigher.