397results about How to "Implement classification" patented technology

The invention provides a bridge structure crack position identification method based on deep learning and computer vision, and the method comprises the steps: obtaining a multi-view visible light image and a depth image, fusing the visible light image and the depth image according to registration parameters, and generating a fusion image containing the texture and depth information of a bridge; for the fused image, extracting a crack region by adopting an adaptive threshold segmentation algorithm, and dynamically adjusting the threshold to obtain a crack binary image; according to the crack binary image, calculating three-dimensional coordinates of crack pixel points through a three-dimensional reconstruction algorithm, and performing reconstruction to obtain three-dimensional point cloud data containing crack space position and form information; on the basis of the three-dimensional point cloud data, three-dimensional local feature descriptors are adopted to describe geometric attributes of the surface of the crack, crack fragment parameters are calculated, and a three-dimensional crack model is constructed; based on a deep learning model, a transfer learning strategy is adopted, a self-adaptive transferable bridge crack three-dimensional detection method is formed, and model parameters are updated so that the model parameters can adapt to the characteristics of a target bridge.

The invention relates to the technical field of non-destructive testing, and discloses a non-destructive testing method for enhancing defect detection efficiency through deep learning, which comprises the following steps: step 1, acquiring multi-modal data of a target material, including data acquired through X-ray imaging, ultrasonic detection and infrared thermal imaging; 2, signal denoising processing is carried out on the obtained multi-modal data, signal denoising is carried out through wavelet transformation, and the wavelet transformation comprises the steps that multi-scale decomposition is carried out on signals, and high-frequency noise is removed through a soft threshold method; and step 3, carrying out image enhancement processing on the de-noised multi-modal data. The internal structural features of the material are captured from multiple physical dimensions by fusing multi-modal data of X-ray imaging, ultrasonic detection and infrared thermal imaging. The multi-modal fusion can make up for the deficiency of a single detection technology, so that the defect identification in complex structures and materials is more comprehensive and accurate, thereby significantly improving the detection precision.

The invention relates to the technical field of radar induction, and discloses a danger alarm method, device and equipment based on radar induction and a storage medium, and the method comprises the steps: carrying out the preprocessing of three-dimensional point cloud data collected by a laser radar and original image data collected by a security camera, and obtaining target point cloud data and target image data; constructing a target Voronoi diagram of the monitoring area; performing feature extraction and feature fusion to obtain a fused feature vector; inputting the fusion feature vector into a deep learning model for target detection and classification to obtain an initial detection result of each potential dangerous target; carrying out danger level evaluation to obtain the danger level and confidence of each potential danger target; according to the method, the monitoring parameters of the laser radar and the security and protection camera are optimized, the cooperation monitoring parameter combination of the equipment is output, the monitoring parameters of the laser radar and the security and protection camera are optimized in real time, cooperation monitoring of the equipment is achieved, and the overall monitoring efficiency of the system is improved.

The invention discloses an intelligent software testing method and system based on automatic script generation, and relates to the technical field of software testing, and the method comprises the following steps: deploying a distributed data acquisition probe, dynamically adjusting sampling frequency, collecting software operation data, and extracting data features based on a deep learning model; based on the data features, establishing a behavior pattern classification system, and mapping behavior patterns to realize hierarchical behavior classification; the method comprises the steps of constructing a test scene, generating test data, executing a test task, monitoring a test execution state in real time, calculating a test efficiency index, evaluating and feeding back a test effect, and dynamically adjusting a feature extraction strategy. According to the method, deep learning and reinforcement learning technologies are utilized, accurate recognition and classification of software behavior modes are achieved, and comprehensiveness and effectiveness of testing are guaranteed through multi-dimensional coverage rate analysis and a closed-loop optimization mechanism; and meanwhile, the automation level and the intelligent degree of software testing are remarkably improved, and a powerful guarantee is provided for improving the software quality.

The invention discloses an automatic fine ore sorting method and system based on XRT (X ray tomography) and hyperspectrum, which are used for realizing comprehensive fusion of ore surface and internal information by combining X-ray tomography and hyperspectral imaging technologies so as to improve the recognition precision and the sorting efficiency. XRT is used for extracting the internal structure and density characteristics of the ore, hyperspectral imaging is used for identifying the mineral composition characteristics of the ore surface, and information complementation is achieved. The system adopts a multi-modal convolutional neural network to perform independent feature extraction on XRT and hyperspectral data, and the extracted features are fused through a deep residual network. And the fused features are processed by a cross-modal attention mechanism to ensure that key features of different modals are fully integrated and optimized, and finally high-precision ore classification and sorting are realized. According to the method, the limitation of a single data source on feature extraction is overcome, the ore recognition and sorting precision is greatly improved, and the refined and intelligent development of ore sorting is promoted.

The invention provides an intelligent industrial quality inspection method and system based on AI visual identification, and relates to the technical field of visual identification, and the method comprises the steps: inputting an industrial product image collected at an edge end into a pre-trained defect region detection model, and obtaining a defect region detection result; extracting defect area features according to a defect area detection result, carrying out dimension reduction processing to obtain compressed defect area features, and deciding to unload a defect classification task to a cloud end or directly carry out defect classification at the edge end according to an adaptive task unloading strategy in combination with working data of the edge equipment; if the defect classification task is determined to be unloaded to the cloud, transmitting the compressed defect area features to the cloud through a network, performing defect type identification on the cloud by adopting a pre-trained defect classification model, outputting a classification result of the defect area, returning the classification result to an edge end, and combining a defect area detection result and the classification result of the defect area to obtain the defect classification task. And generating a defect visualization effect picture and a quality inspection report.

The invention provides a sea island reef landform identification and classification method and system and a storage medium, and relates to the technical field of remote sensing data analysis, and the method comprises the steps: obtaining remote sensing data of a sea island reef research area; performing prediction by using the remote sensing data through the water depth inversion model to obtain water depth data of the sea island reef research area; inputting the water depth data and the remote sensing data into a landform identification and classification model, and obtaining a landform classification result of the sea island reef research area according to the output of the landform identification and classification model; wherein the initial classification model is pre-trained by utilizing the preset water depth data and the preset remote sensing data through comparative learning, and the pre-trained initial classification model is trained according to the preset water depth data and the preset remote sensing data to obtain a landform identification classification model. According to the invention, high-reliability intelligent identification and classification of the topography and landform of the sea island reef are realized, and the classification effect of the topography and landform of the sea island reef is improved.

The invention relates to the technical field of data processing, and discloses an ocean current simulation method and system based on a three-dimensional temperature-salinity value. The method comprises the following steps: carrying out line calculation and adaptive mesh refinement processing on a three-dimensional temperature and salt field simulation model to obtain a model operation scheme; performing three-dimensional visualization processing on the three-dimensional temperature and salt field initial data set and the model operation scheme to obtain dynamic three-dimensional temperature and salt field visualization data, and analyzing the dynamic three-dimensional temperature and salt field visualization data to obtain three-dimensional structure data and evolution process data of the key ocean phenomenon; performing data assimilation processing on the multi-source real-time observation data, the three-dimensional structure data and the evolution process data to obtain a fusion data set; and performing ocean current prediction calculation on the fused data set to obtain a three-dimensional ocean current prediction field, and performing uncertainty quantitative analysis on the three-dimensional ocean current prediction field to obtain ocean current simulation probability distribution data. According to the invention, the accuracy and timeliness of ocean current simulation based on the three-dimensional temperature-salinity value are improved.

The invention provides an AI-based pipe gallery fire operation hidden danger automatic identification method and device. The method comprises the steps of collecting data of a pipe gallery fire operation site in real time and preprocessing the data; identifying and comparing the check points by adopting an image detection classification algorithm based on a diffusion model and a large model distillation technology to obtain an identification and comparison result; and calling a pre-constructed LLM large language model, carrying out potential safety hazard identification on the preprocessed multi-modal data and the identification comparison result through a multi-modal input algorithm, and outputting a multi-modal fire operation comprehensive supervision report. A diffusion model and a large model distillation technology are utilized, data augmentation and data preprocessing are provided, image detection and classification of a pipe gallery fire operation site are realized, and real-time monitoring, automatic hidden danger identification and intelligent early warning of the pipe gallery fire operation site can be realized through LLM construction of an overall system framework and multi-modal input algorithm analysis. And the accuracy and efficiency of potential safety hazard identification are improved.

The invention discloses a power system protocol analysis method and system based on association rule analysis and sequence alignment, particularly relates to the field of private protocol analysis, is used for solving the problems of dynamic field feature extraction and boundary recognition, and aims at extracting features of dynamic fields through separation of static fields and the dynamic fields. Then, hidden frequent items and associated features are mined by using an association rule algorithm, so that dynamic fields are displayed in a structured manner in a feature matrix, which becomes the basis of subsequent clustering and inference. The feature matrix is clustered through a DBSCAN algorithm, protocol formats are classified, noise and abnormal data are processed, and a preliminary format is provided for field boundary recognition. According to the method, the Nedleman-Wunsch sequence comparison and information entropy verification technology is combined, the boundary position is accurately identified, the fuzzy field is dynamically adjusted by using the field boundary ambiguity index and the sequence alignment error correction index, and finally, the boundary division precision is enhanced through mutual information correction, so that the accuracy and processing efficiency of protocol analysis are improved, and the method is suitable for popularization and application. And automatic analysis of a complex protocol is realized.

The patent of the present invention discloses an automatic classification and recycling device for kitchen waste, which specifically relates to the technical field of resource recycling. An automatic classification and recycling device for kitchen waste, including a recycling box, a stepper motor is provided on the recycling box, and a first rotating shaft is connected to the stepper motor; a holding plate is distributed on the first rotating shaft, and a first gravity sensor is provided on the first rotating shaft; a first fixed ring is provided in the recycling box, a threaded rod is provided in the first fixed ring, a clamping plate is connected to the threaded rod, and a second gravity sensor is also provided on the clamping plate; a first driven gear is provided on the threaded rod; a second rotating shaft and a partition plate are provided in the recycling box, a first motor is provided on the second rotating shaft, a blade is provided on the second rotating shaft, and a second motor is also provided on the recycling box, and a driving gear is provided on the second motor, and the driving gear is meshed with two first driven gears. The technical solution of the present invention solves the problem that the existing office building garbage cannot be classified and recycled, and can be used for the centralized recycling of kitchen waste.

The invention discloses an event detection and extraction method and system based on entity and trigger word fusion recognition. The method comprises the steps that trigger words in event text data are extracted; mapping the event text to a vector space based on a word embedding model, and constructing an event text classifier to perform similarity measurement on all words of which the part-of-speech is a verb and the trigger word in the event text; constructing an event detection model to obtain an event category in the event text data; identifying a key element entity of the event; constructing a dependency syntax tree to represent a syntactic dependency relationship of the sentences; obtaining the score of the dependency syntax tree; and inputting the preprocessed word vector and the part-of-speech vector representation obtained through supervised learning into a role tagging model for fusion to obtain a vector representation with part-of-speech features, and after training, carrying out conditional random field processing to obtain a final semantic role tag of the word. Accurate event detection based on entities is realized, and high-quality extraction of events is towed.

The invention belongs to but not limited to the technical field of new energy, particularly relates to a battery physical and chemical model parameter identification method and system based on different temperature impedances, aims to improve the accuracy of model simulation, considers contact resistance when establishing a one-dimensional P2D physical and chemical model of a battery, and simultaneously solves the problem that the model parameters are more and difficult to identify. In order to solve the problems that in the prior art, parameters with high sensitivity and medium sensitivity and meeting actual conditions are selected by adopting a parameter sensitivity analysis method, finally, the parameters selected under different working conditions are identified by adopting a grey wolf optimization algorithm, simulation EIS data are compared with actual measurement data, and the simulation accuracy of the model is influenced by excessive parameters. The result shows that the grey wolf optimization algorithm realizes accurate identification of the parameters of the one-dimensional P2D physical and chemical model of the battery. The method disclosed by the invention achieves remarkable results in the aspects of battery modeling and parameter identification.

The invention relates to the technical field of industrial intelligent manufacturing, in particular to a flexible production line appearance defect optical detection method. Comprising the steps of performing image acquisition on semi-finished products and finished products on a production line; carrying out preprocessing and spatial transformation on the collected image; performing feature extraction and clustering analysis on the positive sample image of the to-be-detected product by using unsupervised learning to form a feature library; comparing the feature vector of the to-be-detected image with a feature library, and outputting the defect position and type of the sample; an improved RTDETR model with supervised learning is adopted, a sample judged as a defect is used for training, and the improved RTDETR model adopts a ResNet and Visual Transform network structure; and performing real-time detection and result feedback on the deployment model. According to the technical scheme, the appearance defect detection efficiency and the product quality of products manufactured on the flexible production line can be improved.

The invention relates to the technical field of image data processing, in particular to a dairy product quality detection method based on image processing. The method comprises the following steps: acquiring a surface grayscale image of a dairy product to be detected in a dairy product container; obtaining a gray feature value of a target pixel point in the surface gray image, and determining a first parameter value of the target pixel point; mapping all the pixel points in the surface grayscale image to a target plane rectangular coordinate system by taking the grayscale feature values of the pixel points in the surface grayscale image and the first parameter values as coordinates; determining a second parameter value of the target pixel point; according to the second parameter value of the target pixel point, the position of the target position point in the target plane rectangular coordinate system is adjusted, and dichotomy is carried out on all the adjusted target position points to obtain abnormal position points; and detecting the quality of the to-be-detected dairy product according to the abnormal position point. According to the technical scheme, the detection of the mould possibly existing in the dairy product can be more accurately realized.

The invention discloses a typical adjustable resource optimization method and system for a power grid adjustment scene, and the method comprises the steps: carrying out the step extraction of operation data for the historical operation data of different adjustable resources; setting typical scene response characteristics by considering response characteristic differences of the adjustable resources in different power grid adjustment scenes, constructing an adjustable resource response potential evaluation index system oriented to the power grid adjustment scenes, and calculating various evaluation indexes; the matching degree difference between the multi-type adjustable resource response characteristics and the power grid adjustment demand is considered, an adjustable resource comprehensive evaluation model is constructed, and adjustable resource response potential comprehensive evaluation in a specific scene is achieved; in combination with an adjustable resource response feature combination and a comprehensive evaluation model, rapid grouping and optimization are carried out on adjustable resources by adopting a combination evaluation and affinity propagation algorithm, and classification and optimization of adjustable resource groups with different response potentials are realized; and in combination with classification and optimization results, analyzing user demand response potential, and optimizing energy management and demand response strategies.

The invention discloses a complex pavement damage condition detection method integrated with an unmanned aerial vehicle. The method specifically comprises the following steps: S1, an unmanned aerial vehicle platform and a sensor system; s2, image preprocessing and feature extraction; the invention relates to the technical field of intelligent traffic. According to the complex road surface damage condition detection method and device fused with the unmanned aerial vehicle, the unmanned aerial vehicle is high in inspection speed and wide in coverage range, the detection period is greatly shortened, the detection cost is reduced, and artificial intelligence technologies such as deep learning and machine learning are combined, so that accurate recognition and classification of road surface damage conditions are realized, and the detection precision is improved; a comprehensive detection report and three-dimensional visual analysis are provided, a scientific basis is provided for road maintenance decision, traffic safety management and urban planning, road maintenance is promoted to develop towards the intelligent and precise direction, the road maintenance efficiency and quality are improved, traffic safety is guaranteed, an interactive decision support system is developed, and the development of the road maintenance system is facilitated. And the user operation convenience and the decision-making efficiency are improved.

The invention discloses a semiconductor spray header hole detection method and device and a medium, and the method comprises the steps: placing a semiconductor spray header on a green LED light source workbench, obtaining a surface image of the semiconductor spray header, carrying out the preprocessing of image data, calculating the geometric feature data and texture feature data of each hole region, constructing the feature vector of each hole, and carrying out the calculation of the feature vector of each hole; establishing a supervised learning classification model based on a support vector machine (SVM), training a classifier by using a historical sample, inputting the feature vector data of each hole into the trained classifier to obtain a defect probability, and determining whether the spray header is qualified or not according to a preset probability threshold; according to the invention, by using computer vision data analysis and machine learning data processing technologies, automatic detection and classification of the surface defects of the spray header are realized, accurate data judgment analysis and recording can be carried out on the hole data of the spray header, and the detection accuracy and efficiency are greatly improved.

The embodiment of the invention provides a traffic attack detection method and device and electronic equipment, and the method comprises the steps: determining a first feature vector corresponding to a to-be-detected data stream based on a feature value, corresponding to a target feature, of each data packet in the to-be-detected data stream; inputting the first feature vector into an attack detection model, and determining a first category of the to-be-detected data stream through the attack detection model; wherein the attack detection model comprises a symmetric tree structure, and the first category represents whether the to-be-detected data stream is normal or not; determining a second category of the first data stream through an attack recognition model; wherein the attack recognition model is a deep learning model, the first data stream is an abnormal to-be-detected data stream, and the second category comprises a preset attack type. A two-stage recognition architecture of the attack detection model and the attack recognition model is adopted, the detection speed is increased, the real-time response capability of detection is improved, fine recognition of network attacks is achieved, and the precision and reliability of detection are improved.

The invention discloses a counting method for system terminal equipment in an NATed network based on an optical network, and belongs to the field of data communication. According to the method, Windows system terminal equipment in a home internal network is monitored and counted in real time by utilizing the high-speed transmission capability and the low-delay characteristic of an optical network; the number of the terminal devices is counted based on the change rate of the IP-ID sequence along with the arrival time of the packet, and different terminal devices are accurately distinguished by analyzing the monotonicity of the IP-ID field in the DNS request packet and the difference of the increase change rate of the IP-ID sequence between different devices. According to the invention, the number of terminal devices can be effectively identified, and the insight ability of an operator for user demands is improved, so that a more personalized service and resource allocation scheme can be provided. Meanwhile, a fixed mode algorithm is adopted, consumption of computing resources is reduced, the recognition speed is increased, the method is very suitable for the scene of recognizing the number of real-time terminal devices, and the stability and reliability of the system are further enhanced.

The invention discloses a GIS switch foreign matter detection method based on multispectral image fusion and related equipment, and belongs to the technical field of electrical equipment foreign matter detection, and the method comprises the steps: obtaining a multispectral image of a to-be-detected GIS switch; and inputting the multispectral image into a pre-trained multispectral fusion model to obtain a multispectral feature map, and inputting the multispectral feature map into a pre-trained multispectral image foreign matter detection model to obtain a foreign matter detection result of the to-be-detected GIS switch. The development and automatic identification of the tiny foreign matter material under different light wavelengths are realized by researching the development technology of the tiny foreign matter material under different light wavelengths and combining multispectral fusion and a multispectral image foreign matter detection algorithm, and the accuracy and robustness of detection are improved.

The invention discloses a hydroelectric generating set excitation power unit fault diagnosis method based on Attention-CNN and a fault knowledge base, and the method comprises the steps: combining normal data and fault simulation data, associating with a corresponding type code, and constructing a single-bridge excitation power unit fault knowledge base; dividing waveform data in the knowledge base into a training set and a test set according to a preset proportion, and inputting the waveform of the excitation voltage and the waveform of the AC side voltage as two-dimensional vectors into an Attention-CNN model for training and testing; and excitation voltage and alternating current side voltage waveforms during the operation period of the excitation system are collected in real time, the excitation voltage and alternating current side voltage waveforms are converted into two-dimensional vectors, the two-dimensional vectors are input into the trained Attention-CNN model, and type codes are output. According to the method, the fault diagnosis efficiency and accuracy of the single-bridge excitation system of the small-sized hydroelectric generating set are improved by using a deep learning technology, automatic identification of a complex fault mode is realized, and the operation stability and the fault processing capability of the small-sized hydroelectric generating set are improved.

The invention provides a battery cell tab cutting and inspecting machine, and relates to the technical field of battery cell tab cutting and inspecting machines, the battery cell tab cutting and inspecting machine comprises a transfer device, a turntable device, a cutting device, a shaping device, a CCD detection device, an OCV / IR detection device and a feeding device, the transfer device comprises a feeding linear module supporting seat, a first bottom plate, a discharging conveying supporting seat and a discharging conveying assembly, the two sides of the surface of the first bottom plate are fixedly connected with a discharging conveying supporting seat and a feeding linear module supporting seat correspondingly, a discharging conveying assembly is installed on the surface of the upper end of the discharging conveying supporting seat, and a discharging linear module is arranged on one side of the discharging conveying assembly. According to the invention, the problem that the battery cell is inconvenient to cut due to the fact that the position of the battery cell is easy to deviate in the cutting process of the battery cell tab and the welding process of the protection plate when the battery cell tab needs to be cut and the protection plate needs to be welded on the battery cell in the packaging process of the battery cell is solved.

The invention provides a cellular lung disease condition prediction method based on multi-modal time sequence data, and belongs to the technical field of cellular lung disease condition prediction. The technical problem to be solved is to provide a cellular lung disease condition prediction method based on multi-modal time sequence data. According to the technical scheme, the method comprises the steps that CT description texts and CT image data of a patient at different time points are collected, then the collected data are cleaned and labeled, and a comprehensive time sequence data set is constructed in combination with the data and used for training and testing a model; a time sequence model is built, the time sequence model comprises an input layer, a wavelet transform layer, a Transform layer, a fusion layer and an output layer, a built time sequence data set is input into the input layer, time sequence data is converted into embedded representation, and the input layer processes multi-modal data and prepares to be input into a subsequent layer; the method is used for predicting the condition of the honeycomb lung disease.

The invention provides a method for identifying and classifying reaction products of an alkali-activated fly ash and slag gelling system. The method comprises the following steps: acquiring a BSE image and an EDS image of an alkali-activated fly ash and slag gelling system sample; clustering based on the pixel color features of the EDS image to obtain each phase mask; fusing the phase feature of the BSE image into the EDS image, and extracting a hole structure based on the gray feature of the BSE image in the same view field area as the EDS image; according to the phase masks and the pore structures, obtaining an unreacted precursor phase mask from which the pore masks are deducted and a reaction product phase mask from which the pore masks are deducted; performing superpixel segmentation on the reaction product phase mask to obtain an Mg / Al ratio of each superpixel; and carrying out clustering analysis on the Mg / Al ratio of each superpixel to realize identification and classification of slag reaction products, mixed reaction products and fly ash reaction products. According to the method, refined classification of the reaction products in the fly ash slag alkali-activated gel system can be achieved, and the accuracy of recognition and classification of the reaction products is improved.

The invention discloses a glaucoma fundus image classification method, which comprises the steps of acquiring existing glaucoma fundus image data information, and performing classification labeling and preprocessing to obtain a training data set; constructing a glaucoma fundus image classification initial model based on a ConvNeXt model, and training to obtain a glaucoma fundus image classification model; and adopting the obtained glaucoma fundus image classification model to classify actual glaucoma fundus images. The invention further discloses an imaging method comprising the glaucoma fundus image classification method. According to the method, the glaucoma fundus image classification network with the feature fusion process is constructed through the ConvNeXt model, so that classification of the glaucoma fundus images can be achieved, the method is suitable for classification of monocular images and binocular images, reliability is high, accuracy is good, the application range is wide, and the method is more objective and scientific.

The invention belongs to the technical field of image processing, and particularly relates to a fresh meat quality evaluation method and system based on multi-modal learning, and the method comprises the steps: obtaining a fresh meat image; extracting meat texture density features of the image; performing multi-level grid division on the fresh meat image based on the meat texture density feature to obtain a texture density feature vector and a color feature vector; taking each grid region after grid division as a node, taking the obtained texture density feature vector and color feature vector as node features, and constructing a multi-scale graph neural network comprising a texture feature network and a color feature network; and performing multi-modal learning feature fusion on the node features according to the multi-scale graph neural network to obtain a texture density comprehensive feature vector and a color comprehensive feature vector, calculating the comprehensive feature vector, evaluating the texture complexity and color distribution characteristics of the meat quality region, and completing the quality evaluation of the fresh meat.

The invention discloses a method for predicting brain glioma subtypes through magnetic resonance imaging based on deep learning, and relates to the technical field of chemical exchange saturation transfer magnetic resonance imaging. The method comprises the following steps: preprocessing a CEST magnetic resonance image of a patient; voxel Z spectrums and corresponding classification labels in tumor and opposite side normal tissue areas of the CEST magnetic resonance image are obtained; inputting the obtained voxel Z spectrum and the classification label into a convolutional network model for training; and performing classification prediction on test data by using the trained model. According to the method, a voxel Z-spectrum full-spectrum training convolutional network model is utilized to realize classification of tumor tissues and normal tissues and prediction of IDH and MGMT subtypes of glioma. According to the method, the whole voxel Z spectrum serves as network input, Z spectrum information can be better utilized, and the requirement for the training data size is lowered.

The invention discloses a heterogeneous graph convolution epilepsy classification model construction method, an epilepsy monitoring device, equipment and a medium, and relates to the field of intelligent medical instruments. The method comprises the steps of firstly constructing a feature extraction module; then constructing an adaptive graph convolutional network model fusing the face and the motion, namely an F2AGCN network model; and training an F2AGCN network model by using a modal coordination learning mechanism to obtain a trained F2AGCN network model, and taking the trained F2AGCN network model as a heterogeneous graph convolution epilepsy classification model. According to the method, the adaptive graph convolutional network model fusing the face and motion information is constructed and used for epilepsy classification, a modal coordination learning mechanism is introduced to force model learning to keep feature representation of semantic consistency between skeleton and face modals, it is ensured that fused features keep semantic consistency between different modals, and the classification efficiency of epilepsy classification is improved. Therefore, integration effectiveness is improved, and the constructed heterogeneous graph convolution epilepsy classification model can directly realize epilepsy classification and monitoring based on video data.

The embodiment of the invention provides an anomaly detection method and system for an air conditioning unit, electronic equipment and a storage medium, and belongs to the field of artificial intelligence and the field of predictive maintenance of industrial equipment. The method comprises the steps that operation parameters of the air conditioning unit are collected and input into an anomaly detection model, anomaly time sequence parameters in the operation parameters of the air conditioning unit are output, and the anomaly detection model is constructed based on a TimesNet algorithm; and inputting the abnormal time sequence parameters into a fault classifier to obtain a fault classification result of the air conditioning unit. A time sequence model of normal operation of the unit is established by using a Times Net algorithm, and multi-sensor operation parameters of the air conditioning unit are detected in real time. And inputting the time sequence interval in which the problem is found into a fault classifier which is jointly constructed based on a mode matcher and a multi-classifier to classify the type of the fault. Abnormality detection and classification of the air conditioning unit equipment are achieved, and the equipment operation efficiency is effectively improved.