77results about How to "Strong representative" patented technology

The invention provides a prediction model construction method and system for personalized equipment operation and maintenance, and relates to the technical field of data processing, and the method comprises the steps: collecting equipment operation data and environment data, extracting the change characteristics of the equipment operation state, carrying out the correlation analysis of the environment data based on historical equipment fault records, and obtaining a prediction model. Performing normalization processing, performing time sequence analysis, generating a joint feature matrix, segmenting the joint feature matrix into time slices of a fixed window according to a time sequence, classifying the time slices into a normal state and an abnormal state, performing feature contribution degree analysis, generating a feature weight matrix, and performing feature extraction; re-combining the equipment features in the time slice feature data set with the environment factors according to feature weights, generating fusion feature data layer by layer, performing supervised learning training on the fusion feature data, constructing an equipment prediction model, and outputting the equipment prediction model as a prediction model of personalized equipment operation and maintenance; according to the invention, the operation and maintenance effect of personalized equipment is improved.

The invention discloses an insulator defect detection method based on feature extraction and cross-scale fusion, and belongs to the technical field of insulator defect detection methods. According to the method, a defect detection model is obtained through training to find defects existing in a detected insulator image, and the method comprises the following steps: S1, inputting an image data set and a corresponding label data set; s2, designing a lightweight fast multi-scale extraction module to complete feature extraction; s3, designing an efficient feature fusion module to complete feature fusion; s4, selecting image features from the output of the cross-weighted bidirectional feature pyramid network through intersection-to-parallel ratio perception query selection; and finally, obtaining a confidence coefficient score and a bounding box of the target through a solution terminal, and further finishing decoding. The method effectively solves the problems that an existing detection method is insufficient in feature extraction capacity in the feature extraction stage and large in model parameter in the feature fusion stage.

The invention discloses a human face sensitive data protection method based on an elastic anti-interference privacy protection mechanism, belongs to the field of information processing, and provides a dynamic self-adaptive sampling strategy based on information entropy in a data preprocessing stage, and the sampling rate is adjusted according to the distribution characteristics of a human face data set and the weight of sensitive information. And the model performance is maximized on the premise of ensuring user privacy. In order to obtain a better privacy protection effect, an entropy adjustment differential privacy mechanism is designed, and Gaussian noise of different degrees is adaptively added according to an information entropy value of face data after a face data set is dynamically and adaptively sampled, so that refined privacy protection of face sensitive data under differential privacy constraint is realized. In a model training stage, an elastic anti-interference training method is provided, diversified adversarial samples are dynamically generated in different stages of model training, and the over-fitting risk of the model is reduced while the robustness of the model to various types of adversarial attacks is enhanced.

The invention discloses an alfalfa SNP (Single Nucleotide Polymorphism) molecular marker combination and application thereof. The SNP molecular marker combination is composed of 61351 SNP molecular markers, the physical positions of the 61351 SNP molecular markers are determined based on sequence alignment of a medicago sativa reference genome ZM-4alfalfa gene, the SNP molecular markers are good in site specificity, high in universality and high in polymorphism, medicago sativa germplasms of different sources can be effectively distinguished, and the application prospect is wide. The alfalfa whole-genome liquid-phase gene chip prepared from the SNP molecular marker can be used for alfalfa germplasm resource genetic diversity evaluation, genetic map construction, functional gene positioning, whole-genome association analysis and the like, and provides important tools and technical supports for alfalfa molecular design breeding and variety authenticity identification. The method has important significance on high-quality development of alfalfa seed industry in China.

The invention discloses a distributed photovoltaic meteorological resource real-time monitoring station group site selection method and system, and the method comprises the steps: 1, obtaining photovoltaic historical resource data of a target region, and constructing a historical data set; 2, based on the historical data set, analyzing space-time distribution characteristics of distributed photovoltaic radiation resources in a target area, and dividing clusters with consistent radiation resource changes; 3, respectively selecting an optimal center point of each cluster; and step 4, performing micro-region site selection based on each cluster center point to determine an optimal site establishment position. According to the method, the existing distributed layout condition is considered, the future distributed development planning is also considered, and the method has the characteristics of full coverage of regional monitoring information, high representativeness of monitoring stations, scientific and reasonable site selection and excellent economical efficiency of a monitoring layout scheme.

The invention provides an ultrasonic sound velocity image generation method, ultrasonic equipment and a storage medium. The generation method comprises the following steps: transmitting ultrasonic waves through a probe of ultrasonic equipment, and receiving echoes of the ultrasonic waves to generate echo signals; based on the receiving delay of each preset sound velocity, performing beam forming processing on the echo signal to obtain an enhanced signal corresponding to each preset sound velocity; and based on the pixel value of each first pixel in the respective sound velocity image of each preset sound velocity, determining the pixel value of a second pixel corresponding to the first pixel in the target image to obtain the target image. On one hand, the sound velocity image obtained by the method is high in representativeness, so that the representativeness of the target image can be improved, and the accuracy of the detection result of the detected object can be improved; and on the other hand, the target image can be generated based on different sound velocity images, so that the information breadth of the target image can be improved, and the display flexibility is higher.

The invention provides a method for characterizing the heterogeneity of a millimeter-level shale oil and gas reservoir. The method mainly solves the problem that the reliability and accuracy of a reservoir evaluation result of an existing volume sampling method are low. The method is characterized by comprising the following steps that S1, shale millimeter-level mineral components are recognized and drilled; s2, based on the obtained powdery sample, oil-containing and non-oil-containing tests are carried out; determining occurrence pore size distribution to carry out reservoir characterization through comparison before and after pumping; s3, on the basis of the obtained powdery sample, carrying out oiliness characterization on different lamina / layered components; and S4, carrying out millimeter-level shale oil reservoir heterogeneity characterization on the basis of the reservoir property and oiliness results of the different strata / layered components. According to the method provided by the invention, the identification precision of the shale formation / stratified structure can be enhanced, millimeter-level fine representation of lithology of different formation / strata is realized, and favorable support is provided for shale oil enrichment mechanism, sweet spot prediction and the like.

The invention relates to the technical field of underground engineering tunnel face information extraction, and discloses an information extraction method for an underground engineering tunnel face rock mass structural plane, and the method comprises the steps: extracting the normal vector of each point from the original point cloud data of the underground engineering tunnel face rock mass structural plane, inputting the converted normal vector into a BIRCH algorithm for clustering to obtain a dominant structural plane category; taking the types of the dominant structural surfaces as input, executing a DBSCAN algorithm on the point cloud of each type of dominant structural surfaces, and segmenting the dominant structural surfaces into independent structural surfaces; performing plane fitting on the independent structural plane by using an RANSAC algorithm to obtain a plane fitting equation, and calculating an inclination angle and an inclination direction of the structural plane according to the plane fitting equation; according to the method, point cloud data provides a rich information source for rock mass structure analysis, an independent structural plane is accurately recognized, and tunnel face rock mass structural plane information is reasonably extracted.

The invention provides a tunnel multi-modal information processing and predicting method based on a Transform technology. The method comprises the following steps: S1, acquiring various sensor data of a tunnel and a vehicle; s2, constructing vehicle composite features according to the sensor data; s3, synthesizing the composite features into a corresponding one-dimensional data sequence, performing one-dimensional convolution processing, inputting processing results into a plurality of corresponding multi-layer perceptron to obtain composite feature weights, and obtaining weighted features according to the composite feature weights, adding position information to the convolution processing result of each one-dimensional data sequence, then carrying out cross-modal fusion, and carrying out projection transformation on the cross-modal fusion result to obtain a corresponding single-modal feature; and S4, weighting the single-mode features by using the weighted features, and inputting a weighted result into the constructed Transform model to obtain a tunnel risk prediction result. The method has the advantages of high prediction accuracy and high efficiency.

The invention relates to the technical field of business process management, in particular to a forestry informatization platform construction method based on business process modeling. The method comprises the following steps of obtaining forestry business process data, performing integrity filling and feature extraction, and generating business process feature data; on the basis, performing process mode mining and systematic modeling to generate an initial business process model; performing sub-domain cutting on the model based on the flow demand data to generate a resource management sub-domain and a production management sub-domain, and performing correlation analysis and structure construction on the resource management sub-domain and the production management sub-domain; the method comprises the following steps: constructing a progressive relationship of a sub-domain association network through actual forestry service data; and according to the progressive relationship, carrying out branch management processing on the initial model, and establishing a process informatization management platform. According to the invention, more accurate, more scientific and more efficient forestry informatization platform construction can be realized.

The invention discloses a purity analysis method for crude vinyl chloride in a conversion process, which comprises the following steps: after sampling, carrying out qualitative and quantitative analysis on small molecule components in the crude vinyl chloride, then analyzing acetylene and high-boiling residues in the crude vinyl chloride, and calculating the purity of the crude vinyl chloride according to a deduction method. Compared with the prior art, the aluminum foil sampling bag is adopted for sampling, so that the problem of poor sample representativeness is solved; the content of the high-boiling residues in the crude vinyl chloride can be accurately analyzed by adopting gas chromatography, so that the accuracy and precision of analysis data are improved; according to the method, gas chromatography is adopted to completely replace a dichloroethane absorption method, so that the working intensity of workers is reduced, and the method has the effects of safety, environmental protection and low energy consumption.

The invention relates to the technical field of intelligent early warning, and particularly discloses an offshore wind power cable temperature trend early warning system, which comprises an initial module for acquiring a topological structure diagram of offshore wind power cables in a preset target area, and determining a target number based on the topological structure diagram; the acquisition module is used for determining a first sequence and setting acquisition points at intervals according to the first sequence and the optimal distance; all collection points are determined, target temperatures are obtained on the collection points, and a target curve is drawn; and the analysis module is used for grouping the acquisition points, determining representative curves according to the target curves in the groups, and determining whether the acquisition points are abnormal or not based on the representative curves. Potential safety hazards possibly existing in the offshore wind power cable can be found in advance, and the reliability and safety of an offshore wind power system are improved.

A detection device and method for Dendrobium officinale health-preserving wine, wherein a main body of the detection device is fixed on the upper end of a base, a bottom plate is slidably provided on the upper end of the base, a sample tank is placed on the upper end of the bottom plate, the bottom plate can drive the sample tank to slide to the inside of a feed port, a liquid outlet pipe is fixedly provided at one end of the base, a discharge end of the liquid outlet pipe is located above the sample tank, and starting a second motor can change the degree of inclination of multiple sampling tubes in a Dendrobium officinale health-preserving wine container, so that the liquid inlet ends of the multiple sampling tubes are respectively located at four positions in the Dendrobium officinale health-preserving wine container, thereby performing multi-position sampling of the Dendrobium officinale health-preserving wine solution, making the sample solution more representative, and thereby improving the accuracy of subsequent detection results of the Dendrobium officinale health-preserving wine.

The invention relates to the technical field of mobile communication, in particular to an urban resident lifestyle identification method based on Wi-Fi handshake data. Wi-Fi handshake data of the mobile terminal are collected; based on the Wi-Fi handshake data, identifying visit events of an individual at different points of interest (POI), and forming a continuous visit sequence according to a time sequence; compressing the visited sequence, and extracting a long-term repeated high-frequency sequence mode; calculating the similarity between the users based on a high-frequency sequence mode, and obtaining a user group cluster with similar lifestyles through clustering; and performing feature analysis on each user group cluster, extracting a core visit sequence, and generating a lifestyle group portrait. Through the technical combination of Wi-Fi handshake data acquisition, sequence compression clustering and sample representativeness correction, comprehensive optimization of urban resident lifestyle identification on coverage integrity, result accuracy, implementation efficiency and cost control is realized, and high-quality data support is provided for urban related decisions.

The invention discloses a large-scale power system high-dimensional probability security scene reduction method and system, and the method comprises the steps: carrying out the sorting and combination of component sets of a multi-dimensional random variable X describing an original scene set according to columns, constructing a distribution matrix V, and calculating a corresponding marginal probability distribution matrix P; dividing the distribution matrix V according to columns and calculating a Cartesian product, obtaining a recombination matrix # imgabs0 # and a marginal probability distribution matrix # imgabs1 #, carrying out descending order arrangement and sampling on the recombination matrix # imgabs2 #, obtaining a compact scene set # imgabs3 #, calculating the order of magnitude of the probability corresponding to each scene in the compact scene set # imgabs4 #, carrying out scene classification, and calculating the probability of each scene in the compact scene set # imgabs4 #; and a scene set # imgabs5 # of each order of magnitude is obtained, an optimization model is constructed by taking minimization of a Wasserstein distance as a target, adaptive selection is performed on the scene set # imgabs6 #, and a union set is taken from the selected sets to obtain an optimal reduced scene set. According to the invention, the scene reduction of any dimension scene can be realized, the representativeness of the reduced scene and the complete coverage of the probability space are ensured, and the application value is extremely high.

This invention provides an intelligent evaluation method for rockburst hazard at working faces based on principal component analysis-probabilistic neural networks, belonging to the field of rockburst hazard evaluation technology. An evaluation index system is established and data is collected by combining rockburst hazard influencing factors, drill cuttings monitoring, critical stress index monitoring, and actual field conditions. Principal component analysis (PCA) is used to simplify the original evaluation index data, resulting in comprehensive evaluation index data containing information from the original evaluation indicators. The comprehensive evaluation index data is divided into a training set and a test set. A probabilistic neural network (PNN) is used to train the evaluation model on the data in the training set, and then the performance of the evaluation model is tested using data from the test set, and its accuracy is calculated. If the accuracy is greater than or equal to 90%, the evaluation model is considered acceptable; if the accuracy is less than 90%, it is considered unacceptable, and the PNN's smoothing factor needs to be modified and retrained until the accuracy of the evaluation result meets the set value.

The invention discloses an annular sampling method for titanium ingot GDMS detection, and belongs to the technical field of titanium ingot GDMS detection sampling. An annular sampling method for titanium ingot GDMS detection comprises the following steps that firstly, a detected titanium ingot is fixed, and surface treatment is conducted on the outer portion of the detected titanium ingot; step 2, determining a turning area: in the turning area 1, feeding leftwards from a position 1 at the speed of 3mm / min, meanwhile, feeding inwards at the speed of 0.025 mm / r, moving to a position 2 after feeding inwards by 2mm, stopping feeding inwards, and horizontally turning leftwards; the titanium ingot can be sampled in an annular sampling mode, the titanium ingot can be sampled in four positions, annular turning is carried out at the four sampling positions, the damage influence of the sampling on the titanium ingot is small, the sampling duration is 15 hours (including titanium ingot clamping), the sampling loss is 10 kg, sampling can be carried out at the 1 / 5 position, the 2 / 5 position, the 3 / 5 position and the 4 / 5 position of the titanium ingot, the representativeness is high, and the sampling accuracy is high. Therefore, the sampling loss of the titanium ingot can be reduced, the sampling time is shortened, and a representative sampling position can be selected.

The present application provides a method for determining a seed-rapid germination type karst plant, and relates to the technical field of vegetation restoration. The method for determining a seed-rapid germination type karst plant comprises: analyzing the configuration category of the embryo and the morphological characteristics of the endosperm of the seeds of the karst plant to be determined. If the configuration category of the embryo of the seed meets the target configuration subtype I and the endosperm is missing or degenerated, the karst plant to be determined is determined to be a rapid germination type; otherwise, the karst plant to be determined is determined to be a non-rapid germination type; the target configuration subtype I includes any one of a spoon type, a straight type, a lateral base type, a stacked type, a peripheral type or a filled type. The method for determining a seed-rapid germination type karst plant provided in the present application adopts a simple and easy biological analysis method, and can accurately and universally determine whether it is a rapid germination type plant by analyzing and determining the configuration category of the plant seed embryo and the morphological characteristics of the endosperm.

The utility model belongs to the technical field of mineral processing metering inspection equipment, and particularly relates to a sampling device of a large flotation machine. Comprising an air delivery pipe, a pulp diversion main pipe, airflow baffles, pulp diversion branch pipes and a sampling storage tank, wherein the air delivery pipe goes deep into the pulp diversion main pipe, the airflow baffle is fixed at the bottom of the air delivery pipe through the connecting rod, the pulp diversion branch pipe communicated with the interior of the pulp diversion main pipe is welded on the side surface of the top of the pulp diversion main pipe, and the sampling storage tank is placed below the pulp diversion branch pipe; according to the utility model, representative ore pulp samples in the large flotation machine can be collected, the pollution problem in the sample collection process is avoided, and the accuracy of the samples is effectively ensured.

The present invention discloses a cerebrospinal fluid tumor organoid culture method, which uses primary tumor cells in cerebrospinal fluid for in vitro tumor organoid culture, mixes the primary tumor cells with Matrigel matrix glue, incubates the mixture to solidify the matrix glue, then adds cerebrospinal fluid tumor organoid culture medium for culture, replaces fresh culture medium every two days, and can induce differentiation to form cerebrospinal fluid tumor organoids after 5 days of culture.

The invention provides a sika deer whole-genome high-density SNP (Single Nucleotide Polymorphism) breeding chip and application thereof, the chip has 100,000 SNP loci in total, and the SNP loci are positioned in a sika deer reference genome (MHLv1.0). According to the SNP locus combination provided by the invention, sika deer genetic typing can be realized through a targeted capture sequencing technology, and results can be used for genetic background analysis, variety identification, germplasm resource evaluation, genetic relationship identification, whole genome association analysis, molecular marker-assisted breeding, whole genome selective breeding and the like of sika deer; and the target SNP sites can be adjusted by directly increasing or decreasing the probes, and compared with a solid-phase chip, the method has better flexibility. Besides, the high-density 100K SNP breeding chip can break through the key technology evaluation restricting sika deer genome selection breeding, lays a foundation for building a sika deer genome selection sharing platform, accelerates the cross development of sika deer breeding from phenotype selection to genome selection, and meets the application of sika deer whole genome selection.

The invention relates to the field of health data analysis, in particular to a human body health data comprehensive analysis system which comprises a video frame type determination module, a frame extraction mode determination module, a data processing module and a data processing module, and is used for determining a video frame type according to an action domain variation degree and a same-frame reference value and determining a frame extraction mode according to frame extraction conditions, the feature acquisition unit is used for acquiring data features of the monitoring data acquired by the changed limb points in the extracted video frames; the mode judging unit is used for determining a monitoring motion state according to the same feature difference reference value and the feature dimension reference value and determining a feature processing mode according to the monitoring motion state; the first processing unit is used for determining a first feature selection mode according to a set condition; the second processing unit is used for determining a feature category according to the individual influence value and the behavior correlation value and determining a second feature selection mode according to a judgment condition; a motion recognition unit; according to the invention, effective features can be screened to improve the behavior recognition precision, so that the motion risk is reduced.

This utility model relates to an automatic sampling device for flotation cell slurry, solving the problems of traditional sampling methods that require repeated sampling, are cumbersome to operate, and have low sampling efficiency. It includes a sampling box, a sampling handle, and a suction pipe. The sampling box has a movable cover on top; the upper half of the sampling box is equipped with grids at certain intervals and angles to allow the slurry to enter the sampling spoon from the side along the gaps between the blades; the lower half of the sampling box has intact side walls without openings; the bottom of the sampling box is inclined at a certain angle to the horizontal plane, and a tubular discharge hole is provided at the lowest point of the sampling box. The lower end of the sampling handle is connected to the sampling box, and the sampling handle is engraved with scales to determine the sampling depth. The suction pipe is connected at one end to the discharge hole at the bottom of the sampling box, and the other end is equipped with a pipe switch. This utility model provides accurate and representative sampling; it has a high degree of automation and high sampling efficiency; the device has a simple structure and is easy to operate.

The utility model belongs to the field of mine exploration and production, and particularly relates to a device for quickly determining the grade of copper-nickel sulfide ore, which is based on rock powder generated in the rock drilling process according to the mine production process and the structural characteristics of the ore structure. The grade of 140-200-mesh ore powder is analyzed through the steps of rock drilling liquid collection, centrifugal classification, drying and cake making, grinding and screening, grade determination through a handheld spectrograph analyzer and the like, and the centrifugal classification device suitable for the content of the utility model is designed according to self requirements. The ore grade determination time is greatly shortened, the problem that errors are large when a handheld spectrograph is used for analyzing mineralized non-uniform ore and star-like-sponge meteorite-like ore is solved, and mine production can be effectively guided.

The invention discloses a feature fuzzy model construction method for infectious disease patients, which comprises the following steps of: 1, constructing a comprehensive patient feature set by collecting medical records, examination reports, medical histories and living habit information of the patients; 2, calculating representativeness, uniqueness and noise immunity scores of each patient, and then selecting patient representatives with characterization capabilities; 3, a fuzzy interval type-2 granularity model containing fuzzy feature descriptions of various patients is constructed around the high total score representation; according to the method, the patient group model which is better in anti-noise capability, higher in interpretability and more in information content can be quickly constructed under the condition that only a small amount of disease related data exists, so that modeling can be quickly performed on the susceptible crowd, fuzzy characteristics influencing disease infection are described, and doctors are assisted in knowing infectious disease characteristics and susceptible crowd characteristics.

This invention discloses an improved near-infrared spectral wavelength selection and modeling method for artificial bee colony algorithms, comprising the following steps: data preparation, chaotic initialization, fitness evaluation, establishing a PLS model based on wavelength subsets, using the regularized fitness function of RMSECV as the fitness Fitnessxi of nectar source xi; searching for foraging bees; searching for observation bees; searching for scout bees; outputting the optimal wavelength subset Sbest and establishing the final PLS model. The beneficial effects of this invention are: the method effectively improves convergence speed and accuracy; chaotic initialization improves the quality of the initial population, avoiding the algorithm from prematurely falling into local extrema; the adaptive perturbation factor balances the algorithm's global search and local exploitation, improving the ability to find the optimal wavelength combination in high-dimensional discrete space, and the selected wavelength subset has stronger representativeness, resulting in a final PLS model with lower RMSEP (root mean square error of prediction) and better robustness.

The invention relates to the technical field of soil sampling, in particular to a portable forest soil multi-point sampling device and a sampling method thereof.The portable forest soil multi-point sampling device comprises a holding mechanism, a mounting mechanism, a sampling mechanism, a sampling pipe fitting and a sealing piece; the holding mechanism is provided with a handheld cross rod for people to hold and use, and a driving cylinder is mounted in the holding mechanism to control the sampling mechanism to work; the mounting mechanism is mounted below the holding mechanism, and a sampling opening is formed in the mounting mechanism and used for soil sampling; the sampling mechanism is mounted below the mounting mechanism; the sampling mechanism is controlled by the holding mechanism to realize digging and sampling work; the sampling pipe fitting is mounted on the mounting mechanism, and is used for collecting and storing the collected soil sample for detection; the sealing element is mounted on the sampling pipe fitting, the sampling pipe fitting is taken out from the mounting mechanism after being collected, and the sealing element is mounted on the sampling pipe fitting for sealing.

This invention discloses a satellite fire detection method based on dynamic exponential features and deep learning. The method includes: acquiring multi-band spatiotemporal observation data from geostationary meteorological satellites; filtering candidate pixels from the full-disk satellite data using multiple threshold conditions; extracting spatiotemporal input features and channel input features of the candidate pixels based on the multi-band spatiotemporal observation data from the geostationary meteorological satellites; and inputting the spatiotemporal input features and channel input features of the candidate pixels into a pre-trained fire spatiotemporal network model to obtain the fire detection result. This invention effectively solves the problem of high false alarms and false negatives caused by cloud cover, high-temperature ground surfaces, and vegetation interference in traditional methods, overcomes the difficulties of weak fire signal extraction and spatial positioning in mixed pixel backgrounds, and significantly improves detection accuracy, stability, and computational efficiency. It can provide key technical support for forest fire monitoring and emergency decision-making.

The invention relates to the technical field of optical detection, and discloses a method for detecting the collagen content of pigskin. The method establishes an initial optical profile of a pigskin sample before contacting a detection medium. And generating a special detection path aiming at the sample according to the contour, and specifying a moving track and a staying node sampled by the probe. Injecting a medium at a path starting point, and starting a synchronization mechanism to synchronize probe movement with a medium diffusion front edge; when the probe arrives at the termination node, the spectrum sequence of each node on the path is completely recorded, and the sequence comprises the dynamic characteristic absorption information of the collagen. And carrying out cross-node correlation fusion on the spectrum sequence to generate a fusion spectrum characteristic spectrum of the sample. And comparing the map with a pre-stored feature-content mapping model, and outputting a collagen content value. According to the invention, adaptive sampling and dynamic process tracking are realized, and the accuracy and reliability of detection are improved.

The invention relates to a network teaching video knowledge point classification extraction method and device, equipment and a medium, and relates to the field of video processing, the method comprises the following steps: using a preset character recognition model to extract text data of each video key frame in a video key frame sequence to generate a text sequence corresponding to each video key frame; inputting a text sequence corresponding to the video key frame into a knowledge point classification model trained to a convergence state, and generating a corresponding knowledge point label for each video key frame; and based on the timestamps corresponding to the video key frames and the knowledge point tags, cutting the network teaching video into a plurality of video clips, and merging the video clips corresponding to the same knowledge point tag to determine the network teaching video corresponding to each knowledge point tag. According to the invention, the network teaching video corresponding to the required learning and teaching content can be quickly and accurately found, and the learning and teaching efficiency is improved.