232results about How to "Capture comprehensive" patented technology

The invention provides a dynamic health adaptive monitoring method and system using artificial intelligence, and relates to the technical field of health monitoring, and the method comprises the steps: carrying out the noise reduction through a hierarchical diffusion probability model through multi-source heterogeneous health data, extracting preliminary features through combining with an improved attention auto-encoder, learning discriminative features through a contrast learning network, and carrying out the recognition of the discriminative features; inputting the discriminative features into an adaptive wavelet neural network for multi-scale feature decomposition, inputting the discriminative features into a dynamic graph neural network to obtain a health feature tensor, obtaining time sequence features by using a hierarchical memory enhancement network, and generating a health knowledge graph in combination with a dynamic routing capsule network and a causal reasoning network; and inputting the time sequence characteristics and the health knowledge graph into a deep Bayesian network for health risk assessment, and obtaining an adaptive monitoring scheme through a model prediction control network by using a hierarchical reinforcement learning network and combining a multi-agent collaborative optimization system and a deep deterministic strategy gradient network.

The invention provides an inventory optimization method and system based on a deep learning model, and relates to the technical field of deep learning, and the method comprises the steps: obtaining the multi-dimensional data of a commodity, constructing a dynamic convolution kernel through employing a graph neural network, constructing a soft clustering distribution matrix in combination with a multi-layer perceptron, and obtaining a long-term dependence feature fusing a space dimension and a time dimension. And fusing the short-term and long-term features, inputting the fused short-term and long-term features into a bidirectional long-short-term memory network to generate a sales prediction sequence, and calculating a prediction confidence interval to obtain a final sales prediction result. And calculating a safe inventory level based on a sales prediction result, generating a replenishment decision by using a dynamic planning model based on deep reinforcement learning in combination with an inventory cost function, and finally sending the decision to a supply chain execution system. The sales volume is accurately predicted through the deep learning model, the inventory management is optimized, the storage cost, the stockout cost and the overstock cost are reduced, and the supply chain efficiency is effectively improved.

The invention discloses a power load prediction method based on multi-source heterogeneous data feature fusion, system equipment and a medium, and relates to the technical field of power system load prediction, and the method comprises the steps: obtaining a power load prediction data set, and carrying out the preprocessing; performing input feature processing on the static data, the graph data and the time series data based on a multi-layer perceptron, a graph convolutional network and a gating loop unit network, and fusing the processed features based on a feature splicing method to obtain comprehensive feature information; based on full-connection neural network construction, taking the comprehensive features as input, inputting training set data into a model, and performing iterative optimization according to a model effect to adjust model parameters and hyper-parameters; and based on the parameter optimization adjustment model, using the evaluated optimal model to predict the power load in the future time period. The method is superior to a traditional method in precision and stability, can better adapt to complex and changeable load prediction scenes, and has wide application prospects and remarkable actual value.

The invention provides a power transmission and transformation project ecological influence space-time boundary determination method and system based on multi-source remote sensing information deep fusion interpretation, and belongs to the technical field of power transmission and transformation project management. The method comprises the steps of data acquisition, data preprocessing and multi-source data integration, multi-scale space-time analysis, ecological influence index construction, remote sensing ecological index model construction, ecological influence space-time boundary determination, dynamic monitoring and early warning system platform establishment and long-term monitoring and evaluation. By integrating satellite remote sensing, unmanned aerial vehicle aerial photography and ground monitoring data and combining an image fusion algorithm, a principal component analysis method and a random forest algorithm technology, the method is beneficial to promoting accurate evaluation of the influence of the power transmission and transformation project on the ecological environment, visualizes the space-time boundary of the ecological influence, breaks through the limitation of a traditional evaluation method on time and space coverage, and improves the evaluation efficiency of the power transmission and transformation project. And a scientific method system for analyzing and evaluating the ecological influence of the power transmission and transformation project is further improved.

The invention discloses a multi-parameter fusion blood glucose prediction method and system. The method specifically comprises the steps that continuous blood glucose monitoring data, physiological feature data and behavior data of a to-be-tested person are obtained; inputting the blood glucose time sequence data into a trained time sequence prediction model to obtain blood glucose time sequence data; fourier transform is carried out on the multi-dimensional feature vector to obtain a multi-dimensional feature vector; constructing a multi-layer feature fusion structure; based on the multi-layer feature fusion structure, the blood glucose time sequence data are fused, and a blood glucose predicted value is output. According to the technical scheme, the blood glucose prediction accuracy is remarkably improved by accurately capturing the blood glucose change trend and periodicity.

The invention provides a network node classification method based on a multilayer graph neural network model, and the method comprises the steps: constructing a graph according to network nodes, carrying out the preprocessing of initial node features in the graph, and obtaining an initial embedded vector; according to the initial embedding vectors of the node and the neighbor node thereof, updating the node representation by adopting a layered message passing mechanism; according to the initial embedded vectors of the nodes and the neighbor nodes thereof, the graph attention network uses an attention mechanism to calculate weights for the neighbor nodes of each node and then dynamically updates feature vectors of the nodes; a multi-head attention mechanism is adopted to splice and merge the feature vectors of the nodes output by the graph attention network and the updated node representation, and the comprehensive feature vector of each node is obtained; the comprehensive feature vectors are further optimized through graph contrast learning and a node embedding module, and optimized features are obtained; and obtaining predicted node classification according to the optimization features. According to the method, complex features of heterogeneous graph data can be effectively captured, the method has good robustness for an over-smoothing problem, more accurate heterogeneous network representation learning is realized, and the method has advanced performance and robustness in a node classification task.

The invention relates to a rock shear failure whole process prediction method based on acoustic emission monitoring, which comprises the following steps: firstly, collecting a rock sample, processing the rock sample into a plurality of rock test pieces, manufacturing an acoustic emission mold, secondly, installing an acoustic emission test platform, carrying out multiple acoustic emission tests of the rock shear failure whole process by using the acoustic emission test platform, and recording data of the multiple acoustic emission tests, the method comprises the following steps: selecting and calculating calculation parameters of a rock shear failure whole-process prediction model based on experimental data, finally constructing the rock shear failure whole-process prediction model based on an XGBoost algorithm, inputting the calculation parameters of the rock shear failure whole-process prediction method into the model, and outputting a rock shear failure prediction result. According to the method, the stability of the rock mass can be continuously monitored in real time, different stages in the direct shear process of the rock are predicted in a classified mode, and the safety and reliability of rock mass engineering such as tunnel engineering can be improved.

The invention belongs to the field of time sequence anomaly detection, and discloses a time sequence anomaly detection method and system based on frequency enhanced convolution Transform, and the method comprises the steps: carrying out the sequence-level preprocessing of an input sequence, including normalization and block generation; encoding the input sequence by adopting a multi-head self-attention mechanism; extracting interrelationships in the sequence by a full convolution module having a plurality of stacked 1D convolution layers of different kernel sizes; s2 and S3 are repeated, and multi-layer coding is carried out on the sequence; converting the encoded embedding into a frequency signal by using a Fourier transform method; the KL divergence and the absolute error are respectively used for calculating the consistency loss of the time domain and the frequency domain, calculating the point-by-point anomaly score, and identifying whether the point is abnormal or not. Wide experiments on five common data sets show that the method is superior to an existing method in the aspect of abnormal recognition.

The invention relates to the field of chronic disease early warning, and discloses a chronic disease early warning intervention system based on artificial intelligence, and the system comprises a data collection module which is used for collecting multi-modal data, and the multi-modal data comprises gene detection data, proteomics data, microbiomics data, environment data and social economic data; the pre-processing module is used for pre-processing the multi-modal data to obtain basic data, and the pre-processing comprises noise removal, abnormal value and missing value processing and targeted data cleaning and feature extraction operation on the data of different modals; and a conversion fusion module. Through cooperative operation of all the modules, a complete and automatic chronic disease early warning intervention process is formed, an organic whole is formed from data acquisition, risk assessment and intervention measure formulation and implementation, and the chronic disease risk whole process management and control are achieved, so that the chronic disease prevention and treatment efficiency and effect are improved.

The invention proposes a hard negative sampling comparison recommendation method based on a graph generation diffusion model, and the method comprises the following steps: coding a user-article interaction graph through a graph convolutional neural network, obtaining a node embedded vector, and carrying out the inner product of the node embedded vector, and obtaining a recommendation probability; the model parameters of the graph convolutional neural network are obtained through the following training steps: S1, generating a comparison view 1 and a comparison view 2 from the node embedding vector through an inline diffusion model view generator and a parameter topological noise reduction view generator; the inline diffusion model view generator comprises an inline diffusion model, and the parameter topological noise reduction view generator comprises a parameter topological noise reduction network; and S2, carrying out comparative learning on the comparative view 1 and the comparative view 2. According to the method, a signal alignment strategy is combined, a double-view generator comparison learning framework is built, and the problems that the model generation capacity is insufficient and the influence of data noise is large are solved, so that the performance of a recommendation system is improved.

The invention provides an elevator risk early warning method based on local patch and enhanced position coding, which comprises the following steps of: firstly, collecting original data of an elevator of the internet of things, and converting the original data into multivariable time sequence characteristics after cleaning the original data; secondly, performing patch operation on the multivariate time sequence, namely, firstly dividing each input univariate time sequence into patches to serve as input tokens of Transformer, so that local semantic information is reserved, and a longer review window can be obtained under the condition that the number of the input tokens is the same; after relative position coding is carried out on the token, the token is input into a Transform model, and in the model, a self-attention mechanism and a convolution pooling layer are adopted to carry out extraction and intersection on patches, so that the feature relation between the patches is concerned, and meanwhile, the time feature relation in the patches is concerned; performing fusion operation on the global time sequence characteristics and the in-patch time sequence characteristics, and then handing over to a full connection layer for prediction to output a prediction result; during model training, the model is adjusted through an optimization algorithm to improve the performance.

The invention provides an AI analysis method and system based on a straight ethyl mirror image, and relates to the technical field of AI.The AI analysis method comprises the steps that multi-scale features of the straight ethyl mirror image are extracted through a pre-trained deep convolutional neural network, and feature fusion is carried out through a self-adaptive feature weight distribution module; and performing lesion area detection on the fused feature map by using a double-branch detection network, and classifying the detected lesion areas. And performing three-dimensional reconstruction based on a continuous orthoscope image sequence, mapping a detection result into a three-dimensional model, realizing accurate positioning and three-dimensional visualization of a lesion area, and recording three-dimensional coordinate position information of the lesion area. The method can improve the accuracy and efficiency of the image analysis of the straight ethyl mirror, and assists a doctor in diagnosis.

The invention discloses a gas prediction method based on the combination of a whale optimization algorithm and a bidirectional long-short term memory network, and belongs to the technical field of gas prediction of iron and steel enterprises. The method comprises the following steps: S1, reading real-time gas data from an online database, preprocessing the data, and constructing a data set required by an experiment; s2, a training set and a test set are divided, the training set is used for determining model parameters, and the test set is used for judging the model effect; s3, constructing a BiLSTM gas prediction model; s4, parameters of the model obtained in the step S4 are optimized; and S5, loading the model of the optimal parameter combination for training and prediction, and storing a predicted value in a database. According to the method, the global optimal search strategy of the whale optimization algorithm (WOA) and the performance of the bidirectional long-short term memory network (BiLSTM) on time sequence data modeling are fused, and the accuracy and efficiency of gas prediction can be remarkably improved.

The invention provides a software system dependency defect tracking method and system based on a graph database, and relates to the technical field of software, and the method comprises the steps: parsing a source code into an abstract syntax tree, and extracting a dependency; constructing a dependency graph data model; establishing a defect mode feature library; performing pattern matching in the graph data model to identify potential defects; calculating a defect influence range and severity; and generating a defect tracking report, including defect priority ranking, complete call chain construction, defect detailed information recording and optimization suggestion generation. According to the method, the dependency relationship defect in a software system can be effectively identified and tracked, and the software quality is improved.

The invention relates to the technical field of data processing, and discloses a multimodal recommendation method and system based on a spectrogram polynomial filter.The method comprises the steps that user data and project data are input into a PolyRec model, and the PolyRec model comprises a multimodal encoder, a behavior encoder and a multimodal alignment module; encoding and fusing the user data and the project data by using a multi-mode encoder to obtain a first feature; encoding the user data by using a behavior encoder to obtain a second feature; and inputting the first feature and the second feature into a multi-modal alignment module for fusion, and outputting a personalized recommendation result. According to the method, the multi-modal characteristics of the user and the behavior information of different frequencies are integrated, so that the user preferences can be captured more comprehensively; and a frequency signal scaler and a unified spectrogram neural network function are adopted, so that the model can flexibly adapt to the characteristics of different frequency signals, and the robustness and adaptability of the recommendation system are further enhanced.

The invention discloses a multi-agent interpretable text psychological health level assessment method and system, and the method comprises the steps: training an original agent model according to a multi-field data set and a psychological dialogue assessment data set, and obtaining a dialogue reference model and an assessment reference model; according to the two reference models, respectively processing the user question data and the multi-round dialogue data to obtain a multi-dimensional dialogue feedback data set and an evaluation feedback data set; according to the multi-dimensional dialogue feedback data set and the evaluation feedback data set, training the two reference models to obtain a dialogue effect evaluation model and an evaluation effect evaluation model; according to the two evaluation models and the psychological dialogue evaluation data set, training the two reference models to obtain a target dialogue model and a target evaluation model; according to the target dialogue model and the target evaluation model, mental health level evaluation is completed. According to the invention, through the multi-dimensional feedback mechanism training model and the cooperative work of multiple models, the interpretability and confidence of the psychological health assessment system are improved.

The invention provides a code review model training method, device and equipment, and a code review method, device and equipment, and the code review model training method comprises the steps: obtaining a training sample which comprises a sample code and review information of the sample code; a retrieval sub-model in the pre-training code review model is utilized to retrieve the sample code, multiple pieces of context information are obtained, and different pieces of context information correspond to different retrieval ranges; combining the plurality of pieces of context information with the sample code to obtain a plurality of pieces of prompt information; generating initial review information of the multiple pieces of prompt information by utilizing a generation sub-model in the pre-training code review model; selecting output review information from the multiple pieces of initial review information; and based on the difference between the output review information and the review information of the sample code, optimizing model parameters of the pre-trained code review model to obtain a trained code review model. The context understanding capability of the model can be enhanced, and the understanding of the model on complex codes is improved.

The invention provides a keratoconus image classification method based on an improved deep learning self-attention mechanism, and the method comprises the steps: obtaining a four-refraction report graph through a three-dimensional anterior segment analysis and diagnosis system, carrying out the image preprocessing of the four-refraction report graph through employing a method for extracting a region of interest, and obtaining a four-refraction report graph data set; constructing a network model based on a multi-directional axis attention mechanism, and inputting the four-refraction report graph data set into the constructed network model based on the multi-directional axis attention mechanism for training to obtain a trained network model based on the multi-directional axis attention mechanism; and carrying out image preprocessing and data enhancement on the four-refraction report graph to be processed, and inputting the four-refraction report graph to be processed into the trained network model based on the multi-directional axis attention mechanism to carry out cornea conus classification. According to the method, unique directional feature information in the corneal topographic map is efficiently captured, the information of the whole image can be comprehensively considered, and the diagnosis accuracy is improved.

The invention discloses a double-domain combined narrow-band radar moving target classification and identification method and related products. Processing a target echo signal received by the radar to obtain a distance Doppler spectrum; obtaining a micro-Doppler time-frequency spectrum based on a target echo signal received by a radar; respectively cutting the distance Doppler spectrum and the micro Doppler time-frequency spectrum to obtain distance Doppler domain characteristic data and micro Doppler domain characteristic data; and inputting the distance Doppler domain feature data and the micro Doppler domain feature data into a pre-constructed RFICNet to obtain a target classification result. According to the method, dual-domain data features of the distance Doppler domain and the micro-Doppler domain are combined, the method does not depend on complex echo data preprocessing operation and excessive prior physical assumptions, dual-domain joint feature information can make up for the deficiency of feature basis during classification, and the classification recognition precision of the narrow-band radar moving target is improved.

The invention discloses a time series anomaly detection method based on a neural network, and the method comprises the steps: obtaining to-be-detected multivariable time series data, and inputting the multivariable time series data into a spatial-temporal feature fusion unit, performing time domain information and attribute information analysis on the multivariable time sequence data by a time domain convolutional network and a graph attention influence network in the spatial-temporal feature fusion unit, and obtaining spatial-temporal fusion features through feature fusion; decomposing the multivariable time sequence data by adopting a self-adaptive time sequence decomposition algorithm to obtain a trend sequence and a periodic sequence; inputting the space-time fusion feature, the trend sequence and the periodic sequence into a time sequence reconstruction unit to obtain a decoding feature; performing anomaly detection on the decoding features through an anomaly detection unit to obtain an abnormal state of the multivariable time series data; according to the method, the reconstruction deviation between the potential abnormal data and the reconstruction data can be expanded, and the potential abnormal data in the multi-dimensional time sequence can be rapidly and accurately detected.

The invention discloses a distributed large model training-oriented multi-granularity semantic tree construction system and a distributed large model training-oriented multi-granularity semantic tree construction method. According to an edge-cloud information fusion space, samples and text tags of a cloud server are embedded and collected into public data set information of edge equipment # imgabs0 # for multi-modal data fusion; the model information interaction space is used for embedding samples and text labels of a cloud server and obtaining a global initial cluster center Ol belonging to Rd * l by adopting a hypergraph clustering algorithm; the cloud server constructs a semantic space # imgabs1 # based on the initial cluster center Ol, and the cloud server distributes the initial cluster center Ol and a text embedding # imgabs2 # of the semantic space to each edge device to construct an edge-cloud cooperative interaction space; the cloud server generates a multi-granularity semantic tree; according to the edge-cloud collaborative interaction space construction global granularity generation model, appropriate data granularity is cooperatively generated in a construction system, and the matching requirement of a heterogeneous model is met.

The invention belongs to the technical field of power grid load prediction, and particularly provides a micro-grid load prediction method and system based on big data, and the method comprises the steps: collecting micro-grid multi-source data of a preset service area through employing an Internet of Things sensor, carrying out the denoising cleaning of the micro-grid multi-source data, and obtaining the processed multi-source data; performing data-level fusion on the processed multi-source data to obtain fused data; performing multi-scale decomposition on the fused data based on an empirical mode decomposition method combined with multi-scale permutation entropy to obtain a multi-scale sub-data sequence; based on the multi-scale sub-data sequence, constructing a micro-grid load prediction model by adopting a time convolution and time sequence mode attention mechanism combined algorithm; and on the basis of the micro-grid load prediction model, predicting a micro-grid load to be predicted to obtain a load prediction result. According to the technical scheme, the accuracy, robustness and calculation efficiency of load prediction can be remarkably improved.

The invention discloses a personnel identification method and apparatus, an electronic device and a storage medium. The method comprises the steps of obtaining a face image of a target object and a feature matrix of a comparison object; performing face region extraction on the face image to obtain a face key part of the target object; performing feature extraction on the face image and the face key parts to obtain a feature point matrix; according to the feature point matrix and the feature matrix, sequentially processing based on distance similarity to obtain a similarity score of each target area of the target object and the comparison object; and carrying out integration operation on the similarity scores of all the target areas to obtain target identification scores of the target object and the comparison object. According to the method, through a multi-layer and multi-region feature extraction and similarity calculation method, the recognition accuracy and robustness of the target object and the comparison object are remarkably improved, meanwhile, the recognition efficiency is improved, and the method has good flexibility and expandability and can be widely applied to the technical field of data processing.

The invention discloses a device for evaluating a swallowing function. The device comprises a point cloud data acquisition module, a three-dimensional modeling and silica gel printing module, a sensor module and a processing and evaluating module, wherein the point cloud data acquisition module is used for acquiring point cloud data of the throat part of a patient; the three-dimensional modeling and silica gel printing module comprises a three-dimensional reconstruction unit and a silica gel 3D printing unit, and a silica gel model is made of a biocompatible silica gel material and has flexibility and plasticity; the sensor module is integrated on the silica gel model and comprises an attitude sensor, a surface myoelectricity sensor, a temperature sensor and a microphone; and the processing and evaluation module is used for performing feature extraction and machine learning analysis according to the data acquired by the sensor module, comprehensively evaluating the coordination, motion range and dynamic features of the swallowing action, generating personalized rehabilitation suggestions for dysphagia, and generating a detailed and comprehensive swallowing ability evaluation report at the same time. The method is non-invasive, real-time, dynamic and highly accurate.

The invention belongs to the technical field of automatic control, and particularly relates to a fault diagnosis and automatic control system of a packed bed continuous hydrogenation process device. The system comprises a model building part used for carrying out material balance modeling on a reactor to obtain a material balance model; establishing a reaction kinetics and heat balance coupling model; establishing a pressure field distribution model; establishing a catalyst inactivation model to obtain activity distribution of the catalyst; the state observation part is used for performing tensor decomposition on the fault deviation vector according to the activity degree part and the pressure distribution of the catalyst in combination with the concentration distribution and the temperature distribution of each component to obtain a fault feature tensor; and the diagnosis and automatic control part is used for calculating an abnormal state probability according to the fault feature tensor, generating a compensation controller according to the abnormal state probability, and performing automatic control by using the compensation controller. According to the invention, comprehensive monitoring and accurate control of the packed bed reactor are realized.

The invention discloses a method and system for detecting alternate side grinding of a steel rail, and the method comprises the steps: obtaining a three-dimensional profile of a steel rail profile in an abnormal shaking region of a subway, and generating an alternate side grinding waveform at a side grinding position according to the three-dimensional profile of the steel rail; and according to the alternate side grinding waveform, the occurrence position of the alternate side grinding phenomenon of the steel rail is determined through the side grinding recognition model. The method has remarkable significance in quantitative evaluation of the steel rail alternate side abrasion phenomenon, the measurement accuracy is improved, and rapid and efficient recognition of steel rail alternate side abrasion is achieved through the side abrasion recognition model. In addition, important technical support can be provided for subway operation maintenance and safety management.

The invention relates to the technical field of intelligent traffic, and discloses a deep learning-based expressway traffic flow detection method and system, and the method comprises the steps: obtaining multi-modal traffic data and environment data in real time from each key road section of an expressway, and carrying out the preprocessing of the multi-modal traffic data obtained in real time, including the time-space alignment of the multi-modal traffic data; inputting the preprocessed multi-modal traffic data and environment data into a pre-trained deep fusion network model, and outputting various vehicle parameters for representing expressway traffic flow characteristics; inputting the various vehicle parameters into a pre-trained high-speed traffic flow detection model, and outputting traffic flow grade information; dynamically updating the deep fusion network model at the key road section by adopting a model fusion mode according to the vehicle parameter change condition at each key road section within the preset time; and updating the high-speed traffic flow detection model according to external conditions introduced by the traffic flow grade information in actual use. And the accuracy and real-time performance of high-speed traffic flow detection are improved by combining deep learning.

The invention relates to the technical field of image processing, in particular to a construction method and system of a water area breach live-action three-dimensional model. The method comprises the following steps: acquiring breach area control point position data; performing surveying and mapping reference confirmation on the breach area control point position data to generate breach area control point reference data; performing three-dimensional coordinate conversion through the breach area control point reference data so as to generate breach area plane coordinate data; unmanned aerial vehicle satellite signal receiving is carried out based on the breach area plane coordinate data, and a standard unmanned aerial vehicle satellite signal is obtained; and performing unmanned aerial vehicle aerial image acquisition based on the standard unmanned aerial vehicle satellite signal to obtain an aerial image of the breach area. Through accurate data acquisition, efficient unmanned aerial vehicle image acquisition and processing, dynamic image utilization and optimized point cloud model construction and texture mapping, the accuracy and timeliness of the three-dimensional model are improved.

The present invention provides a method for identifying the fire source point of a charging place fire based on automatic tracking and positioning by machine vision, which relates to the technical field of fire fighting equipment. A thermal imaging image and a visible light image of a target area are simultaneously acquired and preprocessed, and a suspicious area is located by threshold segmentation. Flame thermal features and flame color features are respectively extracted from the suspicious areas of the thermal imaging image and the visible light image; a flame visual recognition coefficient is obtained by analyzing the flame color features; an edge temperature coefficient and an internal temperature coefficient are obtained by mathematically analyzing the flame thermal features; a fire source recognition coefficient is obtained by mathematically analyzing the flame visual recognition coefficient, the edge temperature coefficient and the internal temperature coefficient, and a fire source is identified by comparing the fire source recognition coefficient with a preset threshold.

The invention discloses a non-meteorological target identification method based on meteorological feature image clustering and fuzzy fusion. Firstly, millimeter wave cloud radar meteorological data are preprocessed; and then generating a meteorological data parameter feature image, and converting the two-dimensional meteorological data array into a two-dimensional pseudo-color image. According to the method, distribution characteristics of meteorological targets and non-meteorological targets on a data image are analyzed, a meteorological data parameter characteristic image is divided into different potential target areas by using an image segmentation large model, and parameter characteristics in the potential target areas are extracted; and calculating the average radial texture and inter-library change degree of the reflectivity factor, the regional average value and variance of the radial speed, and the regional average value parameter characteristic statistical information of the speed spectrum width. A fuzzy logic fusion algorithm is adopted to determine a fuzzy membership function of statistical information, a fuzzy rule is set to calculate a fuzzy membership degree of each data point, and a weighted summation method is adopted to perform defuzzification to obtain a judgment value compared with a threshold value. According to the invention, the accuracy of identifying the non-meteorological target is improved.