4194 results about "Supervised learning" patented technology

The invention discloses an ancient building risk prediction management and control method and system based on a large model, and relates to the technical field of ancient building protection. An ancient building risk prediction management and control system based on a large model comprises a multi-source data acquisition module, a multi-modal data fusion module, a large model risk prediction module, a risk assessment and early warning module, a digital twinborn dynamic management and control module and a repair decision module. According to the method, the structure state data, the visual image data and the environment data are fused, so that the state of the historic building is comprehensively perceived; the self-supervised learning pre-trained ViT-GPT model is combined with the LSTM time sequence model, so that the time sequence analysis capability and the multi-modal data processing capability of risk prediction are remarkably improved, the future damage probability and the risk trend of the historic building can be accurately predicted, and the defects of the prior art in long-time span risk prediction are overcome.

The invention relates to a power equipment fault prediction method based on multi-source heterogeneous data and a twinborn model, and the method comprises the following steps: S1, obtaining multi-modal data related to the operation of power equipment, constructing a joint feature encoder through an attention mechanism in deep learning, and carrying out the joint modeling of multi-modal information, and obtaining a joint feature representation; s2, based on joint feature representation, adopting a dynamic twin model of self-supervised learning and fusion reinforcement learning, and dynamically updating the equipment health state; s3, constructing a complete-cycle prediction framework based on Transform, capturing long-term dependency and dynamic change characteristics in equipment operation, and identifying trigger factors of faults in combination with causal relationship analysis; and S4, based on the equipment health state and the fault prediction result, optimizing equipment operation parameters by adopting an intelligent operation and maintenance optimization engine, and dynamically generating an optimal operation and maintenance plan in combination with a genetic algorithm. The reliability and the operation efficiency of the power equipment are remarkably improved.

A system for AI-supported leak detection and localization in water distribution infrastructures, consisting of: a large number of distributed sensor nodes mounted along a water pipe, each sensor node comprising the following: a pressure sensor configured to measure local hydraulic pressure fluctuations within the pipeline with a resolution of at least 0.01 bar; a flow sensor configured to measure the volume flow within the pipeline with an accuracy of at least ±0.5% of the measured value; an acoustic sensor configured to detect vibration signatures caused by leaks in a frequency range between 50 Hz and 20 kHz; an embedded microcontroller with integrated analog-to-digital conversion circuitry for digitizing sensor outputs; a wireless communication module configured to transmit time-synchronized sensor data to a cloud-based processing platform; and a local energy subsystem with a rechargeable battery and an optional circuit for generating photovoltaic energy; the cloud-based processing platform includes an artificial intelligence engine that comprises the following: a data acquisition module configured to receive and decode the transmitted sensor data and to perform time alignment; a supervised learning module that is trained on historical data of flagged leaks and non-leaks to classify incoming sensor patterns; an unsupervised learning module configured to detect anomalies by modeling normal operating baselines of the pipeline; and a topology-aware localization module configured to determine leak coordinates using the topology of the pipeline network, modeling the propagation of hydraulic waves, and estimating the arrival time difference from multi-node acoustic detections; and wherein the system is configured to provide real-time leak alerts and georeferenced visualization via a remote monitoring interface.

The invention relates to the technical field of red tide anomaly detection, in particular to an ocean red tide anomaly detection method and system fusing multi-source remote sensing and a graph neural network. The method comprises the following steps: acquiring remote sensing image data, unmanned aerial vehicle image data and monitoring data of a monitoring point; performing data preprocessing on the acquired remote sensing image data and unmanned aerial vehicle image data; performing feature extraction and feature fusion on the remote sensing image and the unmanned aerial vehicle image to obtain remote sensing feature data; constructing a space-time diagram structure based on the monitoring data of the monitoring points to obtain diagram structure data; based on a cross-modal comparison self-supervised learning mechanism, carrying out consistency representation learning on a remote sensing feature mode and a graph structure feature mode; by introducing multi-source heterogeneous data and fusing a graph neural network modeling means, the limitation of a single data driving method in the aspects of coarse red tide recognition granularity, low space-time precision and the like is effectively broken through, and the meticulous property and global perception ability of red tide feature modeling are remarkably improved.

The invention relates to the technical field of server load management and dynamic resource scheduling, and discloses a server load state evaluation method based on a dynamic evaluation algorithm, which comprises the following steps: data acquisition: acquiring performance index data, log data and network flow data from a multi-modal data source of a server; data preprocessing: performing time alignment, data cleaning, feature extraction and normalization processing on the acquired data; and self-supervised learning modeling: extracting feature representation from the multi-modal data through a self-supervised pre-training task, wherein the self-supervised pre-training task comprises a time sequence covering prediction task. Through multi-modal data fusion and self-supervised learning, the dynamic characteristics of the server load are accurately extracted, real-time load classification, anomaly detection and trend prediction are realized, the evaluation precision and real-time performance are significantly improved, and at the same time, a resource scheduling strategy of rule driving and reinforcement learning is combined, the server resource allocation efficiency is optimized, and the overload risk is reduced.

The invention relates to dialect speech synthesis, in particular to a semi-supervised dialect emotion speech synthesis system based on hybrid experts, which comprises a text analysis module for preprocessing an input dialect text and generating a text representation vector through feature extraction and feature fusion; the mixed expert module is used for acquiring dialect acoustic features, rhythm features, emotional features and general acoustic features; the dynamic routing module is used for realizing intelligent cooperation among experts through a task-aware soft routing algorithm; the semi-supervised learning module is used for training supervised learning by using the marked dialect emotion voice data and training self-supervised learning by using the unmarked dialect emotion voice data; the acoustic parameter generation module is used for integrating the output of each expert to generate a complete acoustic parameter set; the neural vocoder is used for converting the acoustic parameter set into final dialect emotion voice; the method can effectively overcome the defect that the dialect emotional voice is difficult to accurately synthesize under the condition of lack of sample resources.

Provided is an intelligent grading method and system for pulmonary nodules based on multi-modal feature fusion, including: obtaining ROI and VOI of pulmonary nodules based on chest CT examination images and examination reports by utilizing clinical multi-modal data from physical examination population, designing a multi-task feature extraction network based on attention mechanism, to obtain radiomics features and deep image features from the ROI and VOI; designing a cross-modal feature fusion method based on graph representation learning, designing a multi-modal information extraction method, obtaining specific feature representations and graph structures of modalities, and then fusing the feature representations and the graph structures; and proposing an optimization and clinical verification method of pulmonary nodule grading GCN model based on self-supervised learning, to realize fine grading of pulmonary nodule malignancy with slight differences, thereby providing a new approach to design of fine-grained classification algorithms.

The invention discloses a cloud computing storage, calculation and training resource allocation method based on reinforcement learning and dynamic resource scheduling. The method comprises the following steps: S1, establishing a unified resource scheduling framework; s2, constructing a state space and an action space; s3, generating a state representation for predicting future resource demands; s4, initializing a reward function of key performance index weighted comprehensive evaluation; s5, generating an initial storage resource, computing resource and training resource allocation strategy; s6, continuously adjusting a reward function of the reinforcement learning model based on the self-supervised learning feedback signal; s7, dynamically adjusting a distribution scheme of the storage resources, the computing resources and the training resources through quick response of the resource scheduling framework; s8, optimizing resource allocation among a plurality of cloud computing nodes or platforms; s9, automatically identifying different stages of the current task through a self-supervised learning model; and S10, finally generating an optimal allocation scheme. According to the staged resource scheduling method, the condition of overload or insufficiency of resources is effectively avoided, and the task execution time is greatly shortened.

The invention discloses a network security threat research and judgment method, system and device and a storage medium, and the method comprises the following steps: S1, obtaining network traffic, terminal logs, application program interface calling records and threat intelligence data in real time, carrying out the standardized cleaning and format conversion of the data, and building a unified data lake; s2, matching, identifying and determining threats through a preset known threat feature library, constructing a normal behavior baseline by using an unsupervised learning algorithm, and marking suspicious events deviating from the baseline; s3, for the suspicious event marked in the step S2, mining a potential attack path and an attack intention by combining knowledge graph technology associated asset information, a historical attack chain and a homologous IP address; and S4, based on the attack success probability, the influence asset importance and the diffusion speed, calculating a threat level by adopting a fuzzy comprehensive evaluation model, and generating a research and judgment report containing disposal suggestions.

The invention discloses a power equipment fault diagnosis method and system based on a dynamic knowledge graph and large model collaborative reasoning, and the method comprises the steps: achieving the automatic extraction of an entity relationship through a weak supervision entity relationship extraction mechanism in combination with a power field dictionary and a remote supervision technology, and obtaining a weak supervision entity relationship; constructing a time sequence knowledge graph to capture a dynamic evolution rule of the fault propagation chain; structured knowledge graph embedded representation is fused with a large model input layer through a knowledge injection layer, a two-stage reasoning process is generated by adopting a graph retrieval enlarged model, and finally a diagnosis conclusion containing a structured evidence chain is generated. According to the method, the fusion of weak supervised learning and sequential relation modeling is realized, and the automatic extraction and dynamic updating capability of the knowledge in the electric power field is remarkably improved; through a knowledge injection layer and a two-stage joint reasoning mechanism, the structured reasoning advantage of a knowledge graph and the semantic generation capability of a large model are effectively combined, and the diagnosis accuracy, the time sequence reasoning capability and the interpretability are greatly enhanced.

The invention discloses a high-resolution remote sensing image accurate classification system and method based on deep learning multi-modal fusion, and the method comprises the steps: S1, carrying out the preprocessing and fusion optimization of multi-modal data; S1.1, carrying out the standardization and normalization: carrying out the standardization and normalization of remote sensing data of different modals, and eliminating the influence caused by the difference between the different modals, the difference between the resolution, the spectral range and the like; for optical images, contrast may be enhanced by histogram equalization. The method aims at solving the problems of data heterogeneity, calculation efficiency, over-fitting, difficulty in labeling, real-time performance, interpretability and the like in an existing method by adopting multi-modal data optimization preprocessing, deep fusion model design, automatic labeling and semi-supervised learning, a lightweight model and hardware acceleration technology and a strategy for enhancing interpretability. Through the optimization, the system can maintain high classification precision, improve the calculation efficiency, reduce manual intervention, enhance the transparency and generalization ability of the model, and meet the actual application requirements.

The invention discloses a bridge crack identification and automatic evaluation method based on image identification and AI modeling, and the method comprises the following steps: S1, obtaining an original image of a bridge structure surface, and carrying out the image preprocessing; s2, inputting the standardized image into an image recognition model, performing pixel-level segmentation on a crack region in the image, and outputting a crack mask graph; s3, performing feature extraction processing on the crack mask graph, extracting geometric feature parameters of the crack, and constructing a crack feature vector; s4, constructing an evaluation model based on a supervised learning method, and training the evaluation model; and S5, inputting the crack feature vector into an evaluation model, evaluating the structural risk level of the crack, and outputting a structural risk label. According to the method, image recognition and AI modeling are fused, automatic crack recognition and evaluation are achieved, and the method has the advantages of being high in precision, clear in boundary and intelligent in evaluation.

The invention relates to the technical field of artificial intelligence, can be applied to business scenes of robot strategy training, financial science and technology, medical treatment and health and the like, and discloses a strategy optimization method and device based on an interaction track, equipment and a medium. The reinforcement learning agent interacts with the environment to generate an action and state sequence and record an interaction track, an optimal action strategy is generated based on a track optimization strategy, an original empirical data set is further generated, and a pre-training model is optimized through supervised learning to obtain a target strategy model. According to the method, a task-related high-quality trajectory is generated through reinforcement learning agent and environment interaction, original empirical data is extracted on this basis, and a pre-training strategy model is optimized in combination with a supervised learning mechanism, so that the sample utilization efficiency is effectively improved, and the generalization ability and execution robustness of the model under a multi-task condition are enhanced.

Systems and methods for classification model training can use feature representation neighbors for mitigating label training overfitting. The systems and methods disclosed herein can utilize neighbor consistency regularization for training a classification model with and without noisy labels. The systems and methods can include a combined loss function with both a supervised learning loss and a neighbor consistency regularization loss.

The invention belongs to the technical field of artificial intelligence, and discloses a computer network fault detection system based on an artificial intelligence technology, which comprises a data acquisition module, a feature extraction and preprocessing module and a fault positioning and repairing module. According to the invention, the detection link is accurate and comprehensive, the data acquisition module performs multi-source fusion to obtain rich materials, the feature extraction and preprocessing module generates multi-dimensional vectors to capture complex features, and multi-level fault identification performs comprehensive troubleshooting and reduces misjudgment; the diagnosis process is intelligent and efficient, the hybrid fault diagnosis model fuses supervised and unsupervised learning, processes known faults and detects unknown anomalies, and supervised learning branches are optimized to improve performance; the repair work is timely and reliable, the automatic repair scheme generation method has strategy library matching, dynamic adjustment and rollback mechanisms, quick response and flexible repair can be achieved, and the subsequent diagnosis accuracy is improved through repair verification multi-dimensional evaluation and closed-loop feedback.

The invention discloses an electrical fire intelligent identification system based on multi-dimensional sensor fusion. The system comprises the following steps: constructing a reference environment model through multi-sensor scanning, data dimension reduction and intelligent node deployment; multi-sensor time sequence alignment is carried out through edge calculation denoising and dynamic time warping, and high-priority data is processed in real time through a layering mechanism; establishing a fire feature modeling system through LSTM time sequence analysis, mutual information correlation mining and self-supervised learning; through multi-level data fusion, GAN abnormal data generation and fuzzy logic reasoning; through grading alarm, an intelligent fire extinguishing strategy and remote control, full-process coverage from fire detection to emergency response is realized. A fire scene is visualized by means of a three-dimensional thermodynamic diagram, flame dynamic analysis and an augmented reality technology. The system is suitable for fire detection, alarm and response in a complex industrial scene, and can be widely applied to intelligent management of electrical fire.

The invention provides an emergency command converged communication system resource dynamic scheduling method based on artificial intelligence. The method comprises the following steps: S1, acquiring a plurality of sensors and log data of an emergency command system; constructing a dynamic graph neural network optimization communication topology, and selecting an optimal communication path; combining resource prediction and the optimized communication topology to formulate a global optimal resource scheduling strategy, and generating an optimal resource allocation scheme for each task; the resource pool comprises available computing resources and communication bandwidth; dynamically adjusting resource allocation and a communication path in a task execution process, and generating task execution feedback of a corresponding task; the system state is monitored in real time, and resource prediction, topology adjustment and scheduling strategies are optimized based on execution error feedback of task execution feedback. According to the method, efficient and intelligent dynamic resource scheduling is realized in an emergency command converged communication system through resource prediction of self-supervised learning, self-adaptive topological optimization of a variable topological graph neural network and global scheduling of reinforcement learning driving.

Disclosed is a method for establishing a 3D medical image segmentation model based on masked modeling and application thereof includes: establishing a semi-supervised learning network, wherein a student network includes an encoding module for extracting latent features and a segmentation decoder that predicts segmentation results, a teacher network includes an encoding module and a segmentation decoder that are structurally consistent with the student network; training the semi-supervised learning network, wherein during training, two random masking operations are performed on each image, and the image is input to the two networks respectively; optimizing and updating the weight of the student network, and transferring the updated weight to the teacher network, wherein the training loss function includes prototype representation loss, which is used to characterize the difference between the prototypes extracted and generated by the two networks; the student network may further include a reconstruction decoder and an auxiliary segmentation decoder.

The invention provides a hybrid representation learning-based omni-directional motion control method for a quadruped robot in a three-dimensional environment, aiming at solving the problem that an existing motion control method for the quadruped robot is limited by a narrow forward view angle and external perception noise of the robot, so that the omni-directional motion capability in the three-dimensional environment is difficult to obtain. Comprising the steps of designing various input and output contents of a first-stage end-to-end neural network learning framework, designing a training process including mixed representation learning based on a reinforcement learning framework, constructing a specific neural network model of each module, and constructing a quadruped robot simulation motion control training environment. The method has the advantages that network input and output content beneficial to efficiently learning the relation between actual observation of the robot and the environment truth value is provided, and a mixed representation learning method including supervised learning and comparative learning is introduced, so that the robot can effectively deduce environment information beyond the visual field range of the robot in the training process.

The invention provides a cross-domain target detection method and system based on semi-supervised learning in extreme weather, and the method comprises the steps: enabling a student model to learn bounding box regression knowledge of a trained teacher model through distillation, so as to achieve the regression capability in extreme weather; performing instance-level feature adversarial training on a pre-training detection model which is obtained in a distillation stage and has a regression capability in extreme weather by using an intermediate-state image which is generated by a diffusion model and is closer to a target domain and has smaller noise, and determining a new teacher model and a new student model based on the detection model after the adversarial training; in the fine tuning stage, a dynamic category pseudo label screening strategy is utilized to screen pseudo labels generated by a new teacher model, and a screened target domain pseudo label is utilized to train a new student model to obtain a target detection model used for detecting a to-be-detected image in extreme weather. The phenomena of leak detection and error detection in extreme weather scenes in the prior art can be reduced, and the original reasoning speed is kept.

The invention relates to an industrial control system security policy adaptive configuration method and device and a readable storage medium. According to the method, a control instruction stream, a process variable sampling value and link layer message metadata are synchronously collected, an association hypergraph fusing network semantics, control semantics and physical semantics is constructed to represent a system state, and a sampling period is divided into a plurality of time slots with fixed lengths. In each time slot, self-supervised learning is utilized to extract a topology embedded vector group, and a system operation rule is represented. A robustness envelope interval is generated by calculating a statistical distance between a current time slot and a historical baseline, and is used for driving a reinforcement learning model to generate a security policy parameter set. And executing corresponding access control, traffic shaping and integrity verification operations in the next time slot, generating receipt vector feedback model update according to an execution result, outputting a micro-policy patch with a hash signature and a timestamp, and sending the micro-policy patch to the edge security gateway to realize deployment, thereby realizing closed-loop adaptive optimization of the security policy in the dynamic environment.

The invention discloses a gait sub-phase prediction method and device based on deep learning model fusion. The method comprises the following steps: synchronously collecting multi-source gait data of a wearer walking at various speeds; preprocessing the collected multi-source gait data, dividing a time window by adopting a sliding overlapping window technology, and constructing a supervised learning training data set based on double tags; hierarchically fusing CNN, Bi-LSTM and MHA, and constructing a gait sub-phase prediction model; automatically searching an optimal hyper-parameter combination by using a Bayesian optimization algorithm; a stable prediction time (Pstable) index is defined, and model performance is comprehensively evaluated through one-by-one subject rejection cross validation and a speed migration robustness experiment in combination with a classical index; a modular gait sub-phase prediction system architecture is constructed based on a Simulink platform, and a multi-thread buffer mechanism is integrated to process data streams in real time; and solidifying model parameters to an online reasoning module, and dynamically outputting a prediction result.

The invention relates to the technical field of printed circuit board assembly quality detection, and discloses a PCBA anomaly detection method and system based on three-dimensional modeling and AI fusion, and a medium. The method comprises the following steps: acquiring three-dimensional point cloud data of a PCBA board to be detected; generating a reference three-dimensional digital twin model according to a standard PCBA design drawing; carrying out spatial registration on the three-dimensional point cloud data and the reference three-dimensional digital twin model, obtaining the three-dimensional point cloud data, carrying out hierarchical processing on the obtained three-dimensional point cloud data, extracting geometric features of a welding spot region, contour features of an element region and surface features of a substrate region, and carrying out fusion to generate a feature vector group; constructing a generative adversarial model based on a preset semi-supervised learning framework and the normal PCBA sample vector group; and inputting the feature vector group into a generative adversarial model, and examining the abnormal vectors, the corresponding three-dimensional coordinates and the abnormal types in the feature vector group by the generative adversarial model to complete the abnormal detection of the PCBA board. The method is suitable for quality control of a high-density and miniaturized PCBA.

The invention discloses an industrial machine vision recognition and analysis system, relates to the technical field of machine vision recognition, and is used to solve the problem of poor industrial machine vision recognition. The invention combines total variation denoising, contrast-limited adaptive histogram equalization, random geometric transformation and other technologies to enhance image diversity, improve the system's generalization ability for images in different environments, extract image texture and edge features, provide high-quality data for subsequent analysis, use unsupervised learning to extract image potential features and patterns, and locate abnormal areas through template matching, so as to adaptively discover defects in the production process, dynamically update the model based on a small amount of labeled data, optimize feature representation in real time, and adapt to environmental changes. Finally, semi-supervised learning and pseudo-label generation technology are used to expand the training set, improve the generalization ability of the model, maintain efficient learning when labeled data is limited, reduce manual labeling costs, and improve the accuracy of automated quality inspection.

A method based on a multi-task learning model comprises the following steps: acquiring and processing histopathological image data of gastritis through a data preparation and preprocessing step; feature extraction is performed by using a self-supervised learning pre-trained model, and image blocks are coded into high-dimensional feature vectors; and constructing a multi-task learning model, learning feature representation through a full connection layer module and an attention layer module, and outputting a classification result of each task. And carrying out model training and optimization by using an optimizer, adding multitask loss through a loss function, and dynamically adjusting the model performance. The trained model can automatically detect and grade gastritis, atrophy, acute activity, intestinal metaplasia and other pathological indexes, outputs a standardized evaluation result, and assists in pathological diagnosis. According to the method, a multi-task deep learning framework based on self-supervised learning pre-training is constructed, a traditional single-task modeling mode is broken through, deep learning framework design is driven through pathological index association, and accuracy and clinical practicability of non-neoplastic disease assessment of the stomach are remarkably improved.

The invention discloses a tool wear loss prediction method based on physical information self-supervised learning under the condition of few marks, and belongs to the technical field of numerical control machining center tool wear state prediction. Comprising the steps of collecting monitoring signals such as cutting force, vibration and acoustic emission in the machining process; carrying out preprocessing operation of interception and Z score standardization on each sample; a mixed residual convolutional neural network tool wear monitoring model is designed, and information fusion and high-dimensional representation extraction are carried out on multi-source sensor signals; pre-training the tool wear monitoring model by using the auxiliary task I, and guiding the model to extract the invariance feature of the signal; according to the method, under the condition of scarce marked data, high-precision prediction and excellent generalization ability of tool wear are realized, the fluctuation amplitude of prediction errors can be effectively reduced, and the stability of a prediction result is kept. According to the method, a small amount of mark data is effectively utilized in model training, the stability and reliability of tool wear monitoring are ensured, and the requirements for monitoring precision and adaptability in actual production are met.

The invention provides a medical image segmentation method fusing SAM global modeling and U-Net local optimization, and belongs to the technical field of medical image segmentation of image segmentation. The technical problems that the segmentation precision is insufficient and edge details are lost due to the fact that global semantic features and local detail features cannot be considered at the same time in the segmentation process of a traditional method are solved. According to the technical scheme, the method comprises the following steps of 1, preprocessing a medical image; step 2, constructing a segmentation model fusing a visual Transform and a convolutional network; 3, training the segmentation model by using a supervised learning method with a binary cross entropy loss function; and step 4, inputting an image to be segmented into the trained model, and generating a segmentation result of the medical image. The medical image segmentation method has the beneficial effects that the accuracy and robustness of medical image segmentation are effectively improved by combining the global modeling capability of SAM and the local optimization capability of U-Net.

The invention relates to the technical field of battery health management, in particular to a battery fault unsupervised detection method based on diffusion Transform and confidence coefficient calibration, which comprises the following steps: acquiring multi-modal time sequence data in a battery operation process, and performing preprocessing, including data cleaning, normalization processing, alignment and sampling; a diffusion Transform self-supervised learning framework is constructed, and the framework comprises a diffusion process based on a cosine scheduling strategy, multi-scale Transform architecture coding and a cross-modal self-adaptive fusion mechanism. Through the framework, potential space representation is optimized, and the battery state confidence coefficient is calculated; the optimal detection threshold value is dynamically determined by adopting a Bayesian optimization framework, whether the battery state is normal or faulty is judged according to the comparison result of the battery state confidence coefficient and the optimal detection threshold value, dependence on a fault sample label is completely eliminated, and a high-performance fault detection model can be trained only by utilizing normal sample data.

The invention discloses a construction man-machine safety intelligent detection and early warning method based on deep learning. The method comprises the following steps: S1, introducing a wavelet convolution improved YOLOv11 target detection algorithm framework; s2, coordinate extraction of the instance segmentation module; s3, designing a monocular depth estimation network; s4, training the monocular deep network; and S5, converting the two-dimensional coordinates into the three-dimensional coordinates. According to the invention, the real distance between the worker and the equipment is obtained and real-time safety early warning is carried out; obtaining a man-machine actual distance, and judging the accurate position and posture of the machine by combining angle information of the construction machine with edge contour detection; target detection in a complex construction scene is realized by optimizing a network structure; the S2R-DepthNet depth estimation network is used for performing depth estimation on a single RGB image through a self-supervised learning method, the three-dimensional information acquisition process is simplified, and dependence on expensive hardware is eliminated.

The invention is suitable for the technical field of agricultural intellectualization, and provides an intelligent blueberry disease detection method and system based on multi-modal unsupervised learning, and the method comprises the following steps: carrying out the feature extraction and clustering of preprocessed multi-modal data based on an adaptive contrast deep clustering framework, and obtaining a feature extraction result; obtaining a multi-modal preliminary feature and a preliminary clustering result; based on a multi-modal complementary feature fusion mechanism, according to the multi-modal preliminary features and the preliminary clustering result, carrying out adaptive weighted fusion on the multi-modal preliminary features to obtain multi-modal fusion features; performing unsupervised clustering optimization and disease type identification on the multi-modal fusion features to obtain an unsupervised learning model; and performing deployment and incremental learning on the unsupervised learning model, and detecting the blueberry diseases. According to the method, early-stage accurate detection of blueberry diseases is realized through an unsupervised learning algorithm, a new normal form is provided for intelligent accurate management of blueberries, and the disease prevention and control efficiency and industrial economic benefits are effectively improved.