54 results about "Svm regression" patented technology

The invention relates to the technical field of motor state monitoring and fault prediction, and provides a motor residual life analysis method and system based on a support vector machine, and the method comprises the steps: building a multi-dimensional degradation feature set of a motor, and calculating a health index degradation rate based on multi-source sensor data and a failure threshold; constructing a support vector machine regression model of an adaptive kernel function, predicting a health index change track by using the model and real-time data, and generating a residual life evaluation result if a prediction deviation is within an allowable error range; otherwise, starting an incremental learning mechanism to update the training data set, dynamically adjusting kernel function parameters, and recalculating the trajectory; and if the error requirement is still not met, model regularization parameters are optimized in combination with the working condition data until the residual life evaluation result is converged. According to the method, the accuracy and dynamic adaptability of motor residual life prediction can be improved, and the robustness of the model to complex working conditions is enhanced.

The invention discloses a magnetostriction liquid level meter temperature compensation method and system based on a support vector machine. The method comprises the steps that the currently captured original pulse number is obtained through a magnetostriction liquid level meter; the current environment temperature is detected in real time through a high-precision temperature sensor in the magnetostrictive liquid level meter; taking the original pulse number and the current environment temperature as input features, sending the input features to a pre-trained support vector machine regression model (SVM), and outputting a displacement error prediction value in the current temperature environment; and correcting the original measurement result according to the error prediction value to obtain a real pulse measurement value after temperature compensation. The system comprises a time-to-digital conversion chip, a temperature chip temperature measurement module, a support vector machine reasoning module, a pulse compensation module and a distance calculation module. According to the method, nonlinear modeling can be carried out on measurement errors caused by environment temperature changes, so that the liquid level measurement result is dynamically corrected, and the overall robustness and precision of a measurement system are improved.

The invention discloses a wafer polishing laser interference closed-loop pressure control method, and relates to the technical field of semiconductor manufacturing, and the method comprises the following steps: S1, triggering an automatic measurement program after a ceramic plate and a wafer are overturned; s2, driving a laser interference measuring head to perform non-contact scanning on T2 / T4 point positions of the wafer through a six-axis mechanical arm positioning system to obtain thickness distribution data; s3, fitting the thickness distribution data by adopting a Zernike polynomial to generate a three-dimensional distribution diagram, and calculating a pressure correction amount based on a process dynamic model and a pressure prediction model constructed by a support vector machine regression algorithm; s4, detecting a data outlier in real time through an exception handling module, and triggering an equipment protection strategy when the deviation exceeds a set threshold value; and S5, a pressure control instruction is transmitted to the polishing head through the PROFINET bus by means of the PLC. Through non-contact laser interference measurement, the laser interference technology is applied to thickness detection of the polished wafer for the first time, the single measurement time is shorter than 20 seconds, and the detection efficiency is remarkably improved.

The invention discloses a PID (Proportion Integration Differentiation)-based support vector machine regression monitoring model for VOCs (Volatile Organic Compounds) concentration in water-soil-gas, which comprises the following steps of: performing characteristic parameter extraction on a PID signal according to acquired PID signal data; performing principal component extraction on the characteristic parameters by adopting a principal component analysis method; normalizing a training set and a test set in the process of performing VOCs concentration quantitative analysis by using support vector machine regression; selecting a Gaussian radial basis kernel function as a kernel function of SVM regression, and seeking an optimal parameter delta and an optimal penalty factor C of the radial basis kernel function through a genetic algorithm; and carrying out model construction on PID signals under the condition of different concentrations of VOCs by applying a support vector machine regression method on the optimal parameter delta and the optimal penalty factor C. According to the method, the precision of analyzing the concentration of VOCs is improved, so that the prediction effect on the concentration of VOCs is improved, and the monitoring precision of VOCs is improved.

The invention relates to the technical field of battery management, and discloses a battery fault diagnosis method and system for a CCS module, and the method comprises the steps: obtaining a multi-dimensional data sequence of the voltage, current and temperature of a battery pack, carrying out the abnormality detection, and obtaining a potential abnormal time period and a multi-dimensional data subset; performing clustering analysis on the multi-dimensional data subset to obtain an abnormal feature clustering center, and performing dynamic association analysis to obtain an association mode vector set; calculating the similarity of each single data and a fault propagation path, and matching the propagation path with a preset mode to obtain a fault propagation score; fusing the clustering center and the propagation score, positioning a fault monomer through a support vector machine regression model, and determining a fault monomer identifier; and carrying out segmented clustering and logic judgment on the data corresponding to the identifier, and determining a fault type. According to the invention, accurate positioning and type determination of fault monomers can be realized.

The invention discloses a CFD (computational fluid dynamics) and energy consumption simulation-based thermal performance optimization method for a transformer substation enclosure structure, relates to the technical field of transformer substation building energy conservation, and aims at solving the problem of excessive CFD simulation examples when an annual meteorological parameter-near wall surface temperature prediction model is constructed based on annual typical representative meteorological parameter identification of a clustering algorithm. A meteorological parameter-near wall surface temperature prediction model is constructed based on SVM regression, and the coupling problem of hour-level simulation step length between CFD simulation and energy consumption simulation is solved. According to the method, CFD and energy consumption simulation are coupled, and the problem that in the current energy consumption simulation process, the influence of a high heating source on the outer enclosure structure and building energy consumption is not considered is solved. Therefore, the accuracy of evaluation and optimization of the thermal performance of the enclosure structure of the transformer substation can be improved.

The invention provides a video picture and virtual information projection registration and fusion method and system based on a digital twinning scene, and relates to the technical field of digital twinning, and the method comprises the steps: obtaining a virtual three-dimensional coordinate in the digital twinning scene, and a time sequence video stream and an actual three-dimensional coordinate sequence synchronously collected by multiple cameras; then, space pose calculation is carried out through a Perspective-n-Point algorithm, and an initial mapping relation is obtained; performing deviation registration on the relation by using a support vector machine regression model of time sequence perception to obtain an accurate target mapping relation; determining a target fusion region and extracting visual features of the target fusion region; and finally, inputting the visual features and the image features of the virtual information into a random forest model, generating fusion control parameters in combination with a dynamic change index, and then dynamically projecting the virtual information and fusing the virtual information into a video picture, so that the accuracy of virtual and real projection registration and the naturalness of dynamic fusion can be improved.

The invention discloses a qualitative and quantitative analysis method and system for explosives based on terahertz time-domain spectroscopy, sample preparation is carried out by using unified standards, and the method comprises the steps of grinding, weighing and pressing; the method comprises the following steps: measuring a terahertz time-domain spectrogram of an explosive sample by using a terahertz time-domain spectroscopy system, and obtaining a terahertz frequency-domain spectrogram through Fourier transform; data preprocessing algorithms such as data expansion, SG smoothing, multivariate scatter correction (MSC) and competitive adaptive reweighting (CARS) are used for preprocessing the data; and performing classification / regression analysis by using machine learning algorithms such as a support vector machine classification (SVC) model, a support vector machine regression (SVR) model and a one-dimensional convolutional neural network model (1D-CNN). According to the method, the problems of difficulty in constructing a spectrum database, insufficient algorithm robustness, limited practicability and the like in the prior art are solved, and a more efficient and reliable solution is provided for safety detection and prevention and control of explosives.

The application provides a video picture and virtual information projection registration and fusion method and system based on a digital twin scene, relates to the technical field of digital twinning, and obtains a virtual three-dimensional coordinate in a digital twin scene and a time sequence video stream and an actual three-dimensional coordinate sequence synchronously collected by multiple cameras; then, a space pose solution is performed through a Perspective-n-Point algorithm to obtain an initial mapping relationship; subsequently, a time sequence sensing support vector machine regression model is used to perform deviation registration on the relationship to obtain an accurate target mapping relationship; then, a target fusion region is determined and a visual feature thereof is extracted; finally, the visual feature and an image feature of virtual information are input into a random forest model, a fusion control parameter is generated by combining a dynamic change index, and then the virtual information is dynamically projected and fused into the video picture, which can improve the accuracy of virtual-real projection registration and the naturalness of dynamic fusion.

The present application relates to the technical field of complex equipment component repair, in particular to a device component performance and unit body maintenance level relationship mining method capable of effectively improving the use efficiency of an aero-engine, which first carries out expansion processing of repair samples, and then selects a support vector machine regression method which is better in the condition of small sample problems to solve the mapping relationship between the component performance before repair, unit body maintenance level and component performance after repair. Since a component is generally composed of multiple unit bodies, each component has multiple maintenance levels, and the mapping relationship between the component performance before repair, unit body maintenance level and component performance after repair is a many-to-one mapping relationship. In order to improve the accuracy of support vector machine regression, a hybrid kernel function method is used to optimize it, and a particle swarm algorithm is used to optimize the related parameters.

The invention belongs to the technical field of hydrology and water resource and climate prediction crossing, and discloses a medium-and-long-term runoff prediction method fusing a data enhancement technology and a machine learning model, which comprises the following steps: collecting runoff data and climate system index data of a target area, performing factor screening by adopting a replacement accuracy method, the method comprises the following steps: dividing data into a training set and a test set, constructing a time lag factor, carrying out SMOTE data enhancement on the training set, carrying out runoff prediction model construction by using a support vector machine regression model, carrying out parameter optimization by using grid search, calculating an evaluation index to carry out model performance evaluation, and finally predicting and outputting a medium and long-term runoff time sequence according to the model. The method effectively solves the problem of low prediction precision caused by data scarcity and insufficient factor selection in a traditional method. The method overcomes the problems of data scarcity, imbalance and insufficient model generalization ability in medium and long term runoff prediction, and has wide application prospects in the fields of medium and long term runoff prediction and water resource management.

The invention discloses a chlorogenic acid component detection method fused with deep learning, and relates to the technical field of chlorogenic acid. The method comprises the following steps: acquiring a chlorogenic acid hyperspectral image; clustering the chlorogenic acid hyperspectral image through a fuzzy C-means clustering algorithm based on a plurality of subclass centers to obtain a chlorogenic acid hyperspectral cluster, and calculating a chlorogenic acid spectral cluster feature vector; the method comprises the following steps: calculating mutual information of chlorogenic acid hyperspectral cluster feature vectors to obtain hyperspectral cluster feature weights; obtaining a chlorogenic acid hyperspectral cluster weighted vector by combining the chlorogenic acid hyperspectral cluster feature vector and the hyperspectral cluster feature weight; inputting the chlorogenic acid hyperspectral cluster weighted vector into a support vector machine regression model based on high-dimensional multiple scales, and outputting to obtain a chlorogenic acid concentration predicted value; and constructing a chlorogenic acid concentration diagram according to the chlorogenic acid concentration predicted value, calculating a concentration abnormal value through a sliding window method, and performing early warning if the concentration abnormal value is greater than a preset threshold value.

The application discloses a wafer polishing laser interference closed-loop pressure control method, and relates to the technical field of semiconductor manufacturing, and comprises the following steps: S1: a ceramic plate is triggered to automatically measure a program after being turned over with a wafer; S2: a laser interference measurement head is driven by a six-axis mechanical arm positioning system to non-contact scan T2 / T4 points of the wafer, and thickness distribution data is acquired; S3: a Zernike polynomial is adopted to fit the thickness distribution data to generate a three-dimensional distribution graph, and a pressure correction amount is calculated based on a pressure prediction model constructed by a process kinetics model and a support vector machine regression algorithm; S4: an abnormal processing module is used to detect data outliers in real time, and a device protection strategy is triggered when a deviation exceeds a set threshold; and S5: a pressure control instruction is transmitted to a polishing head by a PLC controller through a PROFINET bus. By means of non-contact laser interference measurement, the laser interference technology is applied to wafer thickness detection after polishing for the first time, and single measurement time is less than 20 seconds, so that the detection efficiency is significantly improved.

The invention belongs to the technical field of power equipment insulation state detection, and particularly relates to an XLPE cable water branch quantitative evaluation method and system based on frequency conversion pseudo trapezoidal wave excitation, and the method comprises the steps: applying pseudo trapezoidal wave excitation, and collecting voltage and current signals; wavelet and adaptive filtering are combined for denoising; establishing a Wiener-ANN nonlinear dielectric response model, and fitting cable response through the combination of a dynamic linear neural network module and a static neural network module; obtaining a frequency domain dielectric spectrum and extracting a multi-harmonic characteristic parameter, a loss slope, a hysteresis angle and a time domain nonlinearity; and finally, outputting a water treeing aging comprehensive index WTI through the support vector machine regression model to realize quantitative evaluation of the aging degree. The system comprises a pseudo trapezoidal wave excitation unit, a high-voltage amplification unit, a micro-current acquisition unit, a signal processing unit and an aging evaluation unit. The broadband harmonic response can be obtained through a single test, the test efficiency is improved by more than 75%, the anti-interference performance is high, the evaluation precision is high, and the method is suitable for on-site rapid nondestructive diagnosis.

The invention discloses a strip mine transportation equipment multi-source information fusion monitoring method based on automatic driving, and the method comprises the following steps: 1, collecting a position signal and a vehicle speed signal of transportation equipment and a rotation angle of a steering wheel according to a sampling period; 2, performing data preprocessing on the to-be-fused data of the multi-source information; 3, constructing a support vector machine regression model; 4, establishing a particle swarm optimization algorithm model, and carrying out optimal solution solving on a penalty coefficient and a kernel function parameter of the support vector machine regression model to obtain an optimal parameter combination; and step five, the optimal parameter combination is substituted into the support vector machine regression model, so that the monitoring result is more accurate, and the travel route monitoring result of the transportation equipment is obtained. The method has the characteristics that the operation time is greatly shortened, the monitoring precision can be improved, and the accuracy is improved.

The invention relates to the technical field of software testing, in particular to a software testing data generation method and system based on a multi-chain doliolaria goblaria swarm algorithm, and the method comprises the steps: initializing a population; randomly generating an initial population, and randomly dividing the population into a plurality of sub-chains; performing fitness sorting on individuals in each sub-chain, and dynamically adjusting the proportion of a leader to a follower in the population; updating the position of each sub-chain leader, and obtaining the actual fitness value of the updated leader individual by actually executing an instrumentation program; updating the position of the follower in the sub-chain based on the adaptive inertia factor, and predicting the fitness value of the updated follower individual by using a support vector machine regression model; updating the optimal fitness value and the corresponding position of each sub-chain and the whole population; and judging whether the current iteration reaches the maximum number of iterations or the solving precision requirement. According to the method, the efficiency and coverage rate of software test data generation are remarkably improved, and the calculation overhead is effectively reduced.

This invention discloses a method for optimizing distributed photovoltaic (PV) microgrid systems, belonging to the field of PV microgrid control technology, specifically considering demand-side response and energy storage. The method includes: Step A: establishing a PV day-ahead output prediction model; Step B: establishing an energy storage and load model; Step C: establishing a PV grid integration capacity assessment and optimization model. The day-ahead PV output is predicted using the PSO-SVM regression prediction method. Addressing the severe problem of PV curtailment, the method considers the impact of energy storage and demand response on PV grid integration rate. Starting from constraints such as no node voltage exceeding limits, no power flow overload, and no energy storage exceeding limits, the optimization objective is established by considering the maximum PV grid integration capacity and the minimum network active power loss, forming a distributed PV grid integration capacity assessment and optimization model that comprehensively considers the PV system's grid integration capacity and economic efficiency. Simulation results show that this model can effectively reduce the PV power plant curtailment rate, providing feasibility guidance for the application of PV in the region.

The invention relates to the technical field of tool management and control, and discloses a tool management and control method based on the Internet of Things and an enhanced Apriori algorithm, and the method comprises the steps: collecting tool data through combining the RFID technology of the Internet of Things and a wireless sensor; performing dynamic clustering analysis by using an improved DBSCAN algorithm, screening out a frequent item set by using space-time semantic enhanced Apriori, and generating an association rule; establishing a cost prediction model by using a multivariate adaptive regression spline, constructing a health state prediction model based on an attention mechanism neural network model, and constructing a residual life prediction model by using support vector machine regression; according to the method, the data is deeply mined based on the intelligent algorithm, the association rules are mined in combination with the improved DBSCAN algorithm and the Apriori algorithm fused with the space-time semantics, and decision support is provided for tool management and control. The tool management and control method has the advantages that the data is deeply mined based on the intelligent algorithm, and the association rules are mined in combination with the improved DBSCAN algorithm and the Apriori algorithm fused with the space-time semantics.

The invention discloses a polyurethane ultraviolet aging hardness prediction method based on support vector machine regression, and the method comprises the following steps: collecting hardness data of polyurethane changing with time in an ultraviolet aging process, constructing a data set, and carrying out the preprocessing; establishing a prediction model taking the aging time as input and the hardness value as output based on a support vector machine regression algorithm; the generalization ability of the model is improved through cross validation and hyper-parameter tuning; a plurality of indexes such as mean square error, mean absolute error, decision coefficient and the like are adopted to comprehensively evaluate the model prediction precision; and predicting the hardness of an unknown aging time point by using the trained model to realize reliable estimation of the aging hardness of the polyurethane. The method is high in prediction precision and high in interpretability, and an effective tool is provided for aging performance evaluation and service life prediction of the polyurethane material.

The invention relates to the technical field of bridge engineering, and provides a multi-objective intelligent optimization method for finished cable force of a cable-stayed bridge. The multi-objective intelligent optimization method comprises the following steps: establishing a finite element model of the cable-stayed bridge, and primarily homogenizing the cable force by adopting a minimum bending energy method according to a principle that the cable force of a long cable is large and the cable force of a short cable is small; constructing a support vector machine regression small sample proxy model based on a radial basis kernel function, carrying out cross validation and tuning on hyper-parameters by adopting grid search and a leave-one-out method, and representing a mapping relation between cable force and structural responses such as bending moment and displacement; establishing a multi-objective optimization model taking bending strain energy and resultant displacement as optimization objectives; according to the multi-target intelligent optimization method based on the double-optimal strategy, comprehensive optimal cable force and partial optimal cable force are used as guidance, common cable force is guided to approach to two types of excellent cable force, iterative optimization is adjusted through a direction factor and a strength factor, a Pareto optimal cable force candidate set is updated through non-dominated sorting, hypercube grids and a congestion degree strategy, and the optimal cable force is obtained. And finally obtaining a comprehensive optimal cable force as a finished cable force of the cable-stayed bridge.

The present application belongs to the technical field of photovoltaic power generation, and particularly relates to a photovoltaic power station power generation capacity prediction method based on artificial intelligence. First, relevant power generation influence data is collected, historical power generation capacity data and meteorological data of the collected inverters are combined to obtain inverter power generation capacity data with meteorological data factors, and data cleaning and repair and other pretreatments are performed; a prediction model is constructed by using a multi-scale adaptive gated LSTM, multi-scale input sequences are divided, first power generation capacity prediction data is obtained through LSTM network modeling and attention mechanism feature fusion; a correction coefficient is calculated according to inverter aging and temperature data; the preliminary corrected prediction data is secondarily corrected through support vector machine regression to obtain final prediction data. The method improves data quality, solves the problem of insufficient generalization ability of traditional models, reduces long-term prediction deviation, and improves prediction accuracy and reliability in complex scenarios.

The application discloses a kind of liquor wisdom distillation process parameter optimization method, belong to liquor brewing technical field.The traditional distillation process is dependent on artificial experience, control precision is low, parameter correlation is difficult to quantify, and the problems such as poor optimization efficiency, the steps of the present application are as follows: build full-process industrial internet-of-things platform, collect 20 variable characteristics and 24 base liquor related index characteristics;According to the sampling time, the data is summarized, the abnormal values are removed by principle / box chart, the missing values are filled by random forest to complete data cleaning;Variable and index correlation is mined using algorithm, and modeling variable set is obtained by combining mutual information maximization dimension reduction;Based on support vector machine regression training prediction model, the process optimization parameter manual is generated by analyzing variable contribution degree.The present application can improve the control precision of distillation, quantify parameter correlation, improve the process optimization efficiency, realize standardized production, and is suitable for liquor distillation production guidance.

The present application relates to the technical field of bridge engineering, and provides a cable-stayed bridge completion cable force multi-objective intelligent optimization method: a finite element model of the cable-stayed bridge is established, the minimum bending energy method is adopted, and the cable force is initially uniform according to the principle of "long cable force is large, and short cable force is small"; a small sample support vector machine regression proxy model based on a radial basis kernel function is constructed, grid search and leave-one-out cross-validation are adopted to optimize the super parameters, and the mapping relationship between the cable force and the bending moment, displacement and other structural responses is represented; a multi-objective optimization model with bending strain energy and combined displacement as optimization objectives is established; and a multi-objective intelligent optimization method based on a double optimization strategy is used to guide the general cable force to approach the two types of excellent cable forces, the iteration optimization is adjusted through a direction factor and a strength factor, the Pareto optimal cable force candidate set is updated through a non-dominated sorting, a hypercube grid and a crowding degree strategy, and finally a comprehensive optimal cable force is obtained as the cable-stayed bridge completion cable force.

The application provides a switch cabinet wye contact thermal fault surface temperature measuring method, comprising the following steps: step one, establishing a switch cabinet temperature fluid field simulation model; step two, based on the switch cabinet temperature fluid field simulation model established in step one, conducting streamline analysis of the temperature fluid field to obtain temperature measuring points on the switch cabinet surface corresponding to internal hot spots; step three, obtaining the relationship between the switch cabinet load current, the surface measuring point temperature and the internal wye contact hot spot temperature under different working conditions and defects through orthogonal calculation as an inversion calculation sample, and then based on a support vector machine regression method, completing sample training to form an inversion calculation model, and using the inversion calculation model to realize internal hot spot temperature inversion calculation from the outside. The application can realize temperature measuring point selection at the switch cabinet shell, can realize the combination of various temperature measuring schemes such as temperature sensors, infrared imaging probes and the like, and can monitor the thermal fault of the switch cabinet circuit breaker part without damaging the original structure of the switch cabinet.

The submarine direct-current power supply system fault positioning method based on the multi-element fault current time domain characteristics provided by the application first acquires the current signal in real time by the protection device, judges whether overcurrent exists at the branch unit, marks the current time as the fault time when overcurrent exists, and acquires the current data within 5 ms after the fault by each protection device; then the initial value of the fault current slope, the initial value of the curvature, the initial peak time and the initial peak size at each branch unit are acquired; then the initial value of the fault current slope, the initial value of the curvature, the initial peak time and the initial peak size are input into the support vector machine regression algorithm trained in advance, and the distance y from the fault point to each branch unit is obtained respectively n ; finally, the fault position is determined, the fault position is sent to the shore base station, and the monopolar grounding fault positioning is completed. The application can quickly and accurately distinguish the monopolar grounding fault line and the fault distance of the submarine observation network direct-current power supply system, the application has a small amount of transmission information, and has a low requirement for the transmission information synchronization.

The present invention provides a method for analyzing the crack propagation life of a welded structure based on a support vector machine model. The method comprises the following steps: using hexahedral units to discretize a geometric model of a welded structure, generating a finite element model, and performing static simulation analysis; implanting an initial crack at the center of a weld, and re-dividing a sub-model grid containing the initial crack using tetrahedral units; obtaining a numerical solution of a stress intensity factor at a crack tip through finite element calculation under different values ​​of half-crack lengths, crack positions, and leading edge positions; obtaining an analytical solution using a stress intensity factor calculation formula, and calculating the difference between the numerical solution and the analytical solution; establishing a support vector machine regression model using the half-crack length, crack position, and leading edge position as inputs and the difference as output, and obtaining a stress intensity factor correction formula based on the stress intensity factor calculation formula; substituting the stress intensity factor correction formula into a Paris crack propagation formula to obtain a Paris correction formula, and predicting the fatigue life of the welded structure.

The invention relates to the technical field of target tracking, in particular to a power transmission line hidden danger target tracking method and system based on spectral image processing, and the method comprises the following steps: collecting a multiband spectral image, extracting an initial spectral stable point group, obtaining a spectral reflection peak value and an absorption valley value, constructing a spectral point template, fitting a reflection spectrum curve, and calculating a feature difference. The method comprises the following steps: judging spectral shift through a feature difference, screening abnormal frames, calculating similarity, generating a trajectory matching score, obtaining a score abnormal region, predicting a target position based on a multi-frame velocity vector, and updating trajectory node information. According to the method, abnormal areas are extracted through multiband spectral image analysis, a spectral feature vector set is constructed, a support vector machine regression model is utilized to stabilize a spectral curve, target spectral feature extraction precision is enhanced, spectral offset is calculated, abnormal frames are screened, a matching trajectory is scored, interference is estimated through a velocity vector, and tracking precision is enhanced through a dynamic adjustment mechanism. And the influence of environmental interference on the tracking process is reduced.

The present invention relates to the technical field of intelligent control of rare earth extraction processes, and specifically discloses a dynamic control system for rare earth extraction process control based on data acquisition. The system comprises a real-time monitoring and division module for emulsification risk, a high-risk early warning processing module, and a low-risk area emulsification trend prediction and adaptive control module. Interfacial tension and turbidity data are collected by high-precision sensors, and features are extracted by combining empirical mode decomposition and empirical wavelet transform to construct emulsification trend characteristic values ​​and divide risk areas. A support vector machine regression prediction model is established for low-risk area data to achieve quantitative prediction of emulsification development trends. When the prediction score exceeds a limit, a fuzzy PID algorithm is used to dynamically adjust the stirring rate, phase flow rate, and demulsifier addition amount to achieve closed-loop optimization control of the extraction process.

This invention discloses a method for monitoring salinized arable land based on spectral analysis, comprising the following steps: Step 1: Data preparation: Collect and process data covering the target area; Step 2: Spectral index calculation and fusion: Fuse the salinity index after ground hyperspectral transformation with the salinity index calculated by Sentinel-2 multispectral analysis at the same time through univariate linear regression to generate a hyperspectral-multispectral fusion index dataset and construct a feature variable library; Step 3: Feature variable selection: Select key features from the feature variable library; Step 4: Support vector machine model construction and training: Train the support vector machine regression model; Step 5: Salinity inversion and accuracy verification: Output the predicted soil salinity value, generate a spatial distribution map of soil salinity in the Hetao Irrigation Area based on the predicted soil salinity value, and classify the salinization level of arable land based on the inversion results; Step 6: Dynamic monitoring: Repeat steps up to step 5 to generate a dynamic change map of salinization.

The invention discloses an insulating oil breakdown voltage field detection method, equipment, equipment, a medium and a product, and relates to the field of power equipment detection.The method comprises the steps that a portable terahertz insulating oil breakdown voltage detection platform is built; acquiring time-domain spectral signals of insulating oil samples with different masses by using a portable terahertz insulating oil breakdown voltage detection platform; determining corresponding spectral characteristics according to the time-domain spectral signals of the insulating oil samples with different masses; respectively extracting terahertz characteristic values according to the spectral characteristics of the insulating oil samples with different masses; constructing a sample data set according to the breakdown voltage values of the corresponding insulating oil samples; establishing a breakdown voltage quantitative evaluation model according to the sample data set in combination with a support vector machine regression algorithm; and carrying out breakdown voltage field detection on the to-be-detected insulating oil by using the breakdown voltage quantitative evaluation model. According to the invention, on-site rapid, lossless and accurate detection of the breakdown voltage of the insulating oil can be realized.