21 results about "Support vector regression model" patented technology

This invention provides a method for predicting aging based on metabolite profiles using nonlocal variance enhancement, relating to the fields of medical data analysis and metabolomics. The method acquires LC-MS and / or GC-MS metabolite profile data and the actual age of the subject to be predicted. It performs missing data checks and intra-class normalization on the data, constructs a class-constrained nonlocal variance enhancement feature extraction model, and combines multi-kernel principal component analysis and linear multi-view fusion to obtain fused metabolite profile features. The fused features are then input into a support vector regression model to output predicted age values. The aging degree is determined based on the difference between the predicted and actual age values, which improves the feature expression ability and interpretability of metabolite profile age prediction.

This invention discloses a multimodal interactive interface design method for elderly care robots, relating to the field of intelligent elderly care. The method includes: collecting user facial images, voice signals, and touch operation information to form a raw multimodal data stream; processing the raw multimodal data stream to extract standardized user real-time state feature vectors; inputting the user real-time state feature vectors and historical interaction data into a support vector regression model; obtaining the real-time cognitive load level based on the user's real-time state feature vectors in the current session; and analyzing the rate of change of user touch operation accuracy within a time window using historical interaction data, calculating the acceleration value of capability change and the trend of touch operation accuracy, and fusing them to derive the interaction abstraction level. This invention constructs a joint decision matrix sensitive to the acceleration of capability change, dynamically triggering an adaptive interaction strategy when accelerated capability decline is detected, driving the interface engine to adjust interface rendering and multimodal guidance in real time.

This invention provides a method for predicting the dominant frequency of blasting vibration by integrating parameter fluctuation processing and the Dream Optimization Algorithm (DOA), belonging to the field of underground engineering safety control technology. The system includes a data acquisition and uncertainty modeling module, a data preprocessing and feature construction module, a hyperparameter optimization module, and a prediction model training and result output module. The method obtains relevant parameters through on-site investigation and monitoring, and generates an extended sample set using a combination of probabilistic perturbation modeling and fuzzy triangular modeling. The data is preprocessed and features are selected. The DOA algorithm is used to globally search the hyperparameters of the Support Vector Regression (SVR) model, and the optimal hyperparameter combination is selected by combining a robustness fitness function. The SVR model is trained, and prediction results and uncertainty prediction intervals are generated. This invention can explicitly characterize the uncertainty of geological parameters, achieve efficient global optimization of hyperparameters, improve prediction accuracy, robustness, and generalization ability, and is applicable to different geological conditions and blasting scenarios. It can be extended to various blasting dynamic response prediction tasks.

The application specifically relates to a simulation calculation method for a high-pressure heater split condition of a nuclear power unit, and belongs to the field of performance calculation and simulation of a secondary loop thermodynamic system of a nuclear power unit, and comprises the following steps: a mechanism model for performance calculation of the secondary loop thermodynamic system of the unit is established according to design condition data and normal operation condition data of the unit; a support vector regression model for prediction of key operation parameters of the secondary loop thermodynamic system of the unit is established and trained according to the normal operation condition data and high-pressure heater split condition data of the unit; and the mechanism model and the support vector regression model are fused to perform performance calculation of the secondary loop thermodynamic system of the unit after the high-pressure heater is split. The mechanism model and the support vector regression model are fused, and in the case that actual high-pressure heater split condition data is lacking, the accurate and rapid prediction of key operation states of the nuclear power unit after sudden splitting of one high-pressure heater of the nuclear power unit under normal operation conditions is realized.

The invention discloses a donkey-hide gelatin adulteration quantitative detection method based on a terahertz time-domain spectroscopy technology, and belongs to the technical field of traditional Chinese medicine authenticity identification and quality evaluation.The method comprises the steps that 1, a sample of adulterated donkey-hide gelatin is prepared; (2) collecting spectral data of adulterated donkey-hide gelatin; (3) processing spectrum data of adulterated donkey-hide gelatin; (4) principal component analysis dimension reduction processing; (5) establishing a grid search method, a genetic algorithm and a particle swarm optimization optimization support vector regression model; and (6) training and testing a support vector regression model. According to the method, the adulterated donkey-hide gelatin can be rapidly detected, and an effective detection method is provided for donkey-hide gelatin adulteration. According to the method, the purpose of rapidly, accurately and quantitatively detecting different adulterated colla corii asini can be achieved by utilizing a terahertz spectrum technology which is relatively high in sensitivity and simple in operation compared with a traditional known method.

This invention relates to the field of power device protection technology, and discloses a method and system for transient voltage suppression of silicon carbide power devices. The method includes: acquiring four-dimensional degradation indices of silicon carbide MOSFETs in each surge absorption channel and performing temperature normalization; training a support vector regression model based on a degradation history database to generate a comprehensive health score and remaining lifetime prediction value for each channel; calculating the degradation-corrected efficiency parameters of each channel based on measured degradation index values ​​to generate a full-channel energy state snapshot matrix; solving a multi-channel surge energy allocation scheme using linear programming with a lifetime balancing penalty term based on the energy state snapshot matrix and health state vector; calculating the PWM control parameters for each channel and executing surge energy absorption and feedback. This invention achieves accurate monitoring of multi-channel degradation states and lifetime-balanced energy allocation.

The present application relates to the technical field of flotation froth image recognition, and discloses a fly ash flotation clean coal ash content prediction method and device based on GA-PSO-SVR, the method comprising the following steps: obtaining multiple fly ash flotation froth images and corresponding clean coal ash content values; performing multi-dimensional feature extraction on each froth image, correlating the multi-dimensional features of each froth image with the corresponding clean coal ash content value, and constructing a training data set; using the training data set, based on a support vector regression model, training a clean coal ash content prediction model, wherein the parameters of the support vector regression model are optimized based on a hybrid optimization algorithm composed of a genetic algorithm and a particle swarm optimization algorithm, and the fitness function used by the hybrid optimization algorithm is the weighted sum of the ash content prediction error and the froth image noise weight coefficient. The present application can accurately fit the nonlinear mapping relationship between the froth features and the clean coal ash content, and improve the accuracy and stability of the ash content prediction.

The application belongs to the field of ocean acoustics, and discloses a marine environmental noise modeling method based on multi-stage modeling regression of wind speed, which firstly collects and matches marine noise and wind speed data, and extracts noise power spectrum; then fits the noise spectrum into a gamma distribution, and extracts shape and scale parameters as physical statistical characteristics; further, a two-stage prediction model is constructed: the first stage uses Gaussian process regression to establish a mapping from wind speed to distribution parameters; the second stage uses a hybrid kernel support vector regression model that fuses Wenz empirical physical kernel and RBF data kernel to predict the noise spectrum in the full frequency band. Through the framework of "physical statistical characteristic guidance", the application decomposes complex prediction problems, effectively fuses physical priori and data-driven learning, significantly improves prediction accuracy, generalization ability and model robustness under data scarcity conditions, and provides a high-precision noise prediction tool for marine wind field monitoring and other applications.

This invention relates to the field of teaching evaluation technology, specifically to a teaching quality evaluation method and system based on general education. The method includes the following steps: extracting a sequence of ability achievement and constructing a stable interval; removing outliers to form discrete nodes; calculating course difficulty indicators and difficulty compensation scores; identifying effective continuous scoring periods and generating segmented score averages; and inputting each result into a support vector regression model to generate a quality evaluation result. In this invention, by jointly correcting for abnormal fluctuations in course achievement difficulty and changes in evaluation rhythm, the interference of outlier data on the overall conclusion is reduced, enhancing the comparability between different courses and different stages, improving the stability and interpretability of the evaluation results, making quality judgments closer to the actual teaching situation, and increasing the reference and application value of the evaluation results in teaching management, providing a more targeted basis for teaching feedback and continuous improvement.

The application relates to the technical field of driving parameter estimation, in particular to a road adhesion coefficient estimation method and device, electronic equipment and a storage medium, wherein the method comprises the following steps: collecting state information of a driving vehicle and obtaining a water film thickness; combining a three-degree-of-freedom vehicle model and a Dugoff tire model to obtain a slip rate, a tire side slip angle and a vertical load; adopting a double-layer adaptive cubature Kalman filter to obtain a first estimation value; inputting the water film thickness into a support vector regression model to output a second estimation value; then time-aligning the first estimation value and the second estimation value; calculating a residual error, a double-layer adaptive cubature Kalman filter confidence and a support vector regression model confidence; and obtaining a final road adhesion coefficient. Thus, the problem that related technologies do not effectively fuse the advantages of model-driven and data-driven methods, resulting in low precision, slow response and poor robustness in complex environments when the tire-road adhesion condition suddenly changes is solved.

The application provides a kind of distributed optical fiber microseismic monitoring method and system, it is related to microseismic monitoring technical field;The method is by the three-dimensional optical fiber array of target monitoring area pre-constructed, strain parameter is obtained by Brillouin optical time domain analysis, then polarization modulation measurement is carried out, and microseismic vibration original signal is acquired;The original signal is carried out multi-scale data fusion and filtering target analysis signal, after extracting features, intelligent identification model is used for identification, and microseismic signal information is obtained;The source location is obtained by combining the signal difference of multiple monitoring points and the multi-parameter joint source location, and then the magnitude information is obtained by support vector regression model inversion.The application realizes from signal collection to magnitude evaluation, uses distributed optical fiber technology to overcome the arrangement problem of traditional detector in complex terrain, improves monitoring precision, stability and coverage, realizes no blind area monitoring, provides reliable underground dynamic monitoring scheme for oil and gas exploitation, mine safety and other fields.

This invention relates to the technical field of power equipment operation status monitoring and fault early warning, and in particular to a transformer temperature early warning method and system. The method comprises the following steps: Step 1: Collecting historical transformer temperature data; Step 2: Calculating the sliding window size; Step 3: Training a support vector regression model based on the transformer temperature data collected in Step 1 and the sliding window size obtained in Step 2; Step 4: Training a random forest model based on the transformer temperature data collected in Step 1 and the sliding window size obtained in Step 2; Step 5: Training a gradient boosting regression model based on the transformer temperature data collected in Step 1 and the sliding window size obtained in Step 2; Step 6: Obtaining and calculating the Stacking model; Step 7: Predicting future temperatures; Step 8: Constructing an early warning rule base based on historical data and professional knowledge as the evaluation standard for early warning after transformer temperature prediction; Step 9: Making a judgment and issuing an early warning; this improves the prediction accuracy.

The application discloses an aero-engine fault prediction method, system, device and storage medium, and the method comprises the following steps: obtaining characteristic variables; performing minimum / maximum autocorrelation factor analysis on the characteristic variables by using a MAF method to extract strong autocorrelation factors; constructing a training set by using the extracted strong autocorrelation factors and corresponding residual service life; obtaining a support vector regression model by using the training set; and performing online prediction on the residual service life of the aero-engine by using the support vector regression model. The minimum maximum autocorrelation analysis is introduced for the first time to analyze the state variables of the aero-engine, then a support vector regression (SVR) model is established for the maximum autocorrelation factors by using the support vector regression, and finally the residual life of the current state is predicted by using the support vector regression (SVR) model, so that the accuracy of the residual life prediction is improved.

This invention discloses an intelligent moisture content detection system and method for mine exits based on the natural caving method. The system includes: a mobile inspection platform; an infrared thermal imager and an air-coupled step-frequency continuous wave radar mounted on it; and an industrial control computer and a host computer, each communicatively connected to the other two. The industrial control computer is configured to: control the infrared thermal imager to perform surface scanning of the area to be measured, acquiring surface infrared thermal images; identify and locate suspected high moisture content anomaly areas in real time based on a preset temperature anomaly analysis model; control the mobile inspection platform to move to the area and trigger the air-coupled step-frequency continuous wave radar to perform directional profile scanning; process the radar echo signal and obtain the soil profile moisture content distribution information based on a pre-trained support vector regression (SVR) model; and finally send the results to the host computer for display and storage. This invention, through the cascading of infrared initial screening and radar precision measurement, achieves rapid, accurate, and three-dimensional detection of moisture content in the mine exit area, providing reliable data support for early warning of disasters such as water inrush and landslides.

This invention provides a method, apparatus, device, and storage medium for voice keyword retrieval. The method includes: extracting features and performing cluster analysis on labeled intervals of N keyword sample audios; determining the number of states corresponding to each keyword based on the clustering results; and establishing a network model that includes keyword state sequences and non-keyword absorption states. Features are extracted from the audio to be retrieved, and the extracted features are input into the network model for decoding. Suspected keyword segments are determined based on the probability values ​​of the keyword state sequences in the decoding path. Features are extracted from the suspected keyword segments and input into a pre-trained support vector regression model. The classification score output by the support vector regression model is used to determine whether the suspected keyword segment is a trained keyword. This invention automatically determines the number of states by performing temporal cluster analysis on the features of keyword samples, making it applicable to keyword retrieval for any language and even non-language sounds, thus expanding the scope of application and improving retrieval accuracy.

This invention relates to the field of defect detection technology, specifically to an intelligent detection and control method for seamless steel pipe welds. The method includes the following steps: acquiring material specifications to construct a process feature vector; analyzing image comparison coordinates to output centering deviation values; evaluating energy trends to generate welding control parameters; mapping sound velocity attenuation to construct a defect distribution matrix; and extracting compensation features to update parameters and generate control commands. In this invention, a seamless steel pipe process feature vector is constructed using a support vector regression model. Combined with end edge image analysis, high-precision pipe centering and positioning are achieved. The evolution trend of the composite welding heat source is evaluated to dynamically correct current and laser parameters. Furthermore, based on the ultrasonic sound velocity attenuation characteristics, abnormal frequency bands are accurately selected to construct an internal defect distribution matrix. Frequency and amplitude compensation features are extracted to dynamically update ultrasonic impact parameters. These techniques enable intelligent closed-loop control of the entire seamless steel pipe process, from pre-weld configuration and in-weld control to post-weld inspection and stress relief.

A system for effort estimation in agile software development using multi-stage evolutionary stacking, consisting of: a data acquisition module configured to retrieve software effort records from one or more data set repositories containing historical data from software development projects with characteristics and actual effort values; a data preprocessing module that is operationally connected to the data acquisition module and is configured to receive the aforementioned software effort data sets from the data acquisition module, cleans the received data by removing inconsistencies with missing target values, and normalizes numerical input characteristics to a common range; a first-level ensemble module connected to the data preprocessing module, wherein the first-level ensemble module comprises a variety of heterogeneous basic learners, including a Random Forest model, a Support Vector Regression model, and an Extreme Gradient Boosting model, which generate predictions from each of the heterogeneous basic learners using the preprocessed data sets received from the data preprocessing module; A genetic algorithm optimization module connected to the first layer's ensemble module, configured to: encode weights as a normalized real-valued vector for each of the heterogeneous base learners; apply a fitness function to minimize the mean squared validation error and derive an optimal weight vector; assign optimized weights to the predictions of each of the heterogeneous base learners; and generate weighted predictions based on the optimized weights. a second-level meta-learning module connected to the optimization module of the genetic algorithm, configured to receive the weighted predictions from the optimization module of the genetic algorithm, processes the weighted predictions using a deep multilayer perceptron neural network to learn complex patterns and nonlinear interactions, and generates a final effort estimate for the software; an output processing module connected to the second-level meta-learning module, configured to receive the final effort estimate for the software and process and visualize the data to improve user understanding; and a user interface connected to the output processing module to receive the processed final effort estimate for the software, wherein the user interface is configured to display the processed and visualized final effort estimate for the software.

This invention provides a method, apparatus, equipment, and medium for predicting shale gas content. The shale gas content prediction method includes the following steps: confirming the geological parameters of the target shale and performing grey relational analysis on the geological parameters and shale gas content to identify the main controlling factors of shale gas content; establishing a support vector regression (SVR) model, optimizing the parameters of the SVR model, outputting the optimal parameters, and constructing an optimal SVR model; training the optimal SVR model based on the main controlling factors of shale gas content to obtain the shale gas content prediction model; and predicting the gas content of the target shale based on the shale gas content prediction model. This invention uses grey relational analysis to select the main controlling factors and optimize the support vector regression (SVR) model, thus solving the problem that the performance of the SVR model depends on the selection of its hyperparameters.

This invention relates to the fields of building engineering technology and intelligent monitoring technology, specifically disclosing a BIM-based building construction data management system. The system includes a data acquisition module, a soil stability analysis module, a settlement trend identification module, a foundation condition judgment module, and a risk linkage control module. The system deploys high-precision soil negative pressure sensors and micro-settlement monitoring instruments to collect foundation-related data in real time. It then uses methods such as sliding window mean processing, Daubechies wavelet transform, and empirical mode decomposition to extract soil negative pressure change characteristic values ​​and cumulative settlement trend characteristic values. Furthermore, it integrates these two types of features into a comprehensive feature vector, inputs it into a support vector regression model for foundation risk level assessment, and combines it with the BIM platform to achieve construction progress linkage and hierarchical early warning control.

The application provides a lithium battery residual life estimation method and system, specifically, voltage, current and temperature data of a lithium battery are acquired, and a set of weighted particles representing battery health state are initialized; state prediction and update are performed on each particle, and each particle weight is recalculated according to actual measurement data; when the number of effective particles is lower than a threshold, resampling is performed; the particle set is divided into high-weight and low-weight subsets according to average weight; a support vector regression model is trained with high-weight particle state as input and normalized weight as output; for low-weight particles, K nearest neighbor vectors in the support vector model are identified according to Euclidean distance, a displacement vector is generated based on relative position and Lagrange multiplier weighting, and is superimposed on the original state to generate new particles; the state of the particle set after resampling is weighted and summed to obtain the current health state, and the residual life is estimated.