561 results about "Small sample" patented technology

The invention discloses a small-sample high-density crowd detection method based on text cross-modal guidance, and belongs to the field of computer vision and cross-modal intelligent detection. The method aims at solving the problems that an existing method is high in labeling cost, few in sample, poor in generalization, lack of semantic guidance and insufficient in shielding scene adaptation. According to the method, the semantic guiding capacity of a CLIP text encoder and the feature extraction capacity of DINOv2 and SAM pre-training visual models are fused, three light-weight trainable modules are constructed, and text and visual feature fine-grained alignment is achieved. According to the technical scheme, detection is completed through data acquisition and labeling, feature preprocessing, model initialization, cross-modal feature fusion, mask generation and screening, module optimization and performance iterative optimization. According to the method, high-precision detection can be realized only by a small number of labeled samples, the labeling cost is remarkably reduced, the detection robustness of the high-density shielding scene is improved, and the method has the advantages of flexible scene adaptability and lightweight deployment, and is suitable for scenes such as monitoring video analysis and public safety early warning.

The invention discloses a small sample modeling stability evaluation method and system based on Bootstrap resampling, and particularly relates to the technical field of computers and data intellectualization, the method comprises the following steps: completing data preprocessing and stable stage identification under a unified time base, and forming a segment set; an average block length is used as a decision quantity, an SBB is optimized to construct a sample space, and an optimal block length is determined in combination with variance consistency and nominal coverage rate consistency criteria; carrying out re-sampling training around short / medium / long scales, and collecting layered indexes such as prediction, parameters and features; stability calculation and empirical coverage rate calibration are completed based on intra-scale statistics and inter-scale weighted mixing, and a pseudo-stationary diagnosis score is output; and finally, engineering judgment and backspacing optimization are carried out according to the coverage rate, the correlation maintenance and the risk threshold. The system records random seeds, block metadata and model version generation evidence pointers, has the characteristics of traceability and reverifiability, and is suitable for scenes of production lines, network traffic, finance and the like.

The invention provides an industrial defect sample generation method based on small sample fine tuning and a controllable diffusion model, and the method comprises the steps: employing a surface defect image small sample data set, and screening out a model fine tuning data set through employing a pre-constructed image distribution evaluation module; finely adjusting the weight of the pre-trained stable diffusion model; an industrial defect sample conforming to the specified defect feature is generated by combining a trained space condition control neural network with a preset identifier as a cue word; and calculating a distribution difference value between the industrial defect generation sample and the original small sample data set, selecting images with the distribution difference value smaller than a preset threshold value, and adding the selected images into the original small sample data set to form an expanded training data set. According to the method, the small sample data set is utilized, the diffusion model weight is finely adjusted, the cue word containing the preset identifier, the space condition and the space condition control neural network are combined, the training samples are expanded, the calculation cost is reduced, and the application range is wide.

The invention discloses a grain crop grain classification detection method and system based on small samples, and relates to the technical field of grain quality detection.The method comprises the steps that hyperspectral images of grain crop grains are obtained, quality categories are marked, and then multi-modal features are extracted to construct a classification data set; the data set is used for training a multi-modal fusion classification model, a feature fusion network carries out dynamic weighted fusion on multi-modal features by means of an attention mechanism, a high-dimensional fusion feature vector is generated, and a classifier detects a quality category according to the high-dimensional fusion feature vector. During detection, the multi-modal features of the to-be-detected grains are input into the trained model, and then the quality category can be obtained. The method effectively solves the problems that a single modal method is difficult to deal with complex differences among grain crop grain types, high in hyperspectral data dimension, few in samples, easy to over-fit and the like under the condition of small samples, dynamic focusing of key information is realized by introducing an attention mechanism to optimize modal fusion weight distribution, and the method is suitable for popularization and application. And the classification accuracy and the model robustness are improved.

The invention relates to the technical field of data analysis, and discloses a time series data anomaly detection method and system based on convergence trajectory feature analysis, and the method comprises the steps: collecting and preprocessing multi-dimensional time series data of at least one industrial device belonging to the same type, and obtaining a standardized single device feature sequence of each industrial device; training a baseline auto-encoder model based on the standardized feature tensor to obtain baseline parameters and baseline features of the baseline auto-encoder model; performing disturbance training on the baseline auto-encoder model by taking the standardized single device feature sequence of each industrial device as a single sample based on the baseline parameters to obtain convergence trajectory features of the baseline auto-encoder model, and obtaining a standard digital portrait library based on all industrial devices; and carrying out feature attribution and root cause analysis on the digital portrait. According to the method, the sensitivity and the detection accuracy of weak signals, small samples and complex behavior anomalies are remarkably improved.

The invention provides a production control method and system based on SPC alarm linkage process locking, and relates to the technical field of quality management and control, and the method comprises the steps: deploying an SPC sensing node used for collecting technological parameters at a cutting edge position where a machining tool of production equipment is in contact with a workpiece, or at a contact point where a probe of online measurement equipment is in contact with the surface of the workpiece; the method comprises the following steps: acquiring process parameter data of a key production process in real time through an SPC sensing node deployed at a cutting edge position or a contact point; inputting the process parameter data as real-time data or small sample expansion data into an SPC judgment engine; and performing statistical process control analysis on the real-time data or the small sample expansion data through an SPC judgment engine so as to model data clusters which are densely distributed and have stable parameters in the data into a virtual reference curved surface representing a normal state domain. According to the invention, the accuracy of abnormity determination can be improved, and timely management and control of quality abnormity and minimization of production loss can be realized.

The invention discloses a defect identification model construction method and device based on electric power inspection small samples, a storage medium and computer equipment, and the method comprises the steps: obtaining an electric power inspection picture sample set, and obtaining a defect identification model based on the defect tag carrying condition of each electric power inspection picture sample; dividing the electric power inspection picture sample set into a labeled subset and a non-labeled subset, wherein the number of samples of the labeled subset is significantly smaller than the number of samples of the non-labeled subset; performing fine tuning on the teacher model pre-trained based on the general image data set through the labeled subset to obtain a fine-tuned teacher model; based on the non-label subset, through the fine-tuned teacher model, outputting defect category probability distribution of each electric power inspection picture sample in the non-label subset, and generating a soft label of the electric power inspection picture sample based on the defect category probability distribution; and performing model training on the initial recognition model based on the labeled subset, the unlabeled subset after the soft label is generated and the target loss function to obtain a defect recognition model.

The invention provides a method for constructing a multi-task regression model based on self-supervised pre-training and double-attention element learning, and belongs to the field of artificial intelligence and machine learning. The method comprises the following steps: a data construction and importing stage, a supervision pre-training stage, a dual attention regression model construction stage, a FoMAML-based meta-learning training stage, and realization of rapid adaptation and performance evaluation. Through a four-stage method of self-supervised pre-training + meta-learning optimization + self-attention enhancement + downstream rapid adaptation, high-dimensional structure features are efficiently modeled and accurately predicted under the condition of small samples, the rapid adaptation ability of the model on a new task can be remarkably improved, the local and global feature representation ability is enhanced, and the method has the advantages of being high in robustness and high in robustness. The method gives consideration to calculation efficiency and generalization performance, can also adapt to multi-task high-dimensional complex structure data, and has the advantages of being high in convergence speed, high in adaptability and high in prediction precision.

The embodiment of the invention discloses a terahertz virus spectrum classification method and system based on transfer learning, and the method comprises the steps: obtaining a source domain terahertz spectrum data set containing various biomolecule hydrogen bond vibrations, and obtaining an initial feature extraction model; extracting a hydrogen bond vibration characteristic mode in a preset frequency band range, and taking a characteristic center of the hydrogen bond vibration characteristic mode as a cross-domain bridging characteristic; obtaining target sample data, wherein the sample data is a small sample data set; converting the similarity between the feature center and a target sample spectrum peak into a dynamic weight of a migration loss function, and performing layer-by-layer fine adjustment on the initial feature extraction model network to obtain a virus classification model; the fine tuning process comprises the following steps: adjusting a weight factor in real time according to the spectrum peak offset of the current batch of samples; inputting the terahertz spectrum data of the to-be-detected virus into the virus classification model to obtain a classification result of the to-be-detected virus. According to the application, the accuracy, robustness and industrial applicability of terahertz virus detection are remarkably improved.

The invention discloses a coating quality prediction and optimization method and system based on multi-modal data fusion, and the method comprises the steps: S1, fusing coating parameters, environment data, equipment states and historical quality data, and generating a global feature matrix; s2, inputting the matrix into a GBDT model containing physical prior constraints, and outputting a quality risk probability and a key index; s3, target parameters are generated through a genetic algorithm, Q-Learningg and Pareto screening; s4, performing closed-loop feedback on updated data; s5, cross-line adaptation is achieved through transfer learning and domain adaptation; and S6, iteratively training the model in batches. The system comprises a data fusion module, a quality prediction module, a parameter optimization module, a closed-loop control module, an overline adaptation module and a self-evolution module. The method aims at solving the problems that traditional coating overline parameters are difficult to reuse, small sample modeling cost is high, prediction is separated from process logic, and parameter adjustment lags, so that overline adaptability and prediction stability are improved, small sample cost and energy consumption are reduced, and the method is suitable for high-end coating scenes such as automobiles and aviation.

The invention relates to an equipment cavitation fault detection method based on a CNN and BiLSTM hybrid architecture. The method comprises the following steps: acquiring a time domain vibration signal during equipment operation; performing multi-scale feature extraction on the time domain vibration signal to obtain a three-dimensional feature; and capturing historical laws and future trends of the three-dimensional features, focusing key nodes of fault evolution, and outputting a fault classification probability. The problems of incomplete complex time sequence feature extraction, weak long-range dependence modeling capability and poor small sample scene generalization existing in equipment cavitation fault detection in the prior art can be solved, so that the accuracy and reliability of industrial equipment cavitation fault detection are improved, and the requirement for efficient and stable fault diagnosis in an actual industrial scene is met.

The invention discloses a small sample target detection method for intelligent operation of a robot, and the method comprises the steps: carrying out the preprocessing of a query set, obtaining a standardized query set, carrying out the feature extraction of data in the standardized query set, and obtaining the multi-scale features of the query set, performing feature aggregation and feature conversion on the query set multi-scale features to obtain meta-features; performing feature extraction on the data in the support set according to a multi-scale feature extraction unit to obtain support set multi-scale features; the cross-domain dense feature correlation distillation network re-weights the meta-features and the support set multi-scale features to obtain weighted features; and the multi-scale regression prediction module performs multi-scale regression prediction on the weighted features to obtain target position and category prediction. According to the method, the detection precision of small sample targets in a dynamic unstructured environment is improved, and the problem of precise positioning of uncommon objects in autonomous positioning of the tail end of the mechanical arm is solved.

The invention provides a deep learning-based ultrasonic-assisted CMT welding process parameter optimization method, which comprises the following steps of: acquiring process parameters, ultrasonic parameters, multi-modal sensing data and result quality labels, and constructing a fusion data set containing numerical simulation and actual measurement samples; and introducing physical consistency constraints of energy conservation, mass conservation and temperature-solid phase fraction evolution into a learning target, and establishing a micro-proxy model comprising a spatial-temporal feature encoder, a layer channel adjacency graph network and a physical consistency output head, so as to realize end-to-end prediction from process / ultrasonic parameters and sensing data to multiple quality indexes. According to the uncertainty of the agent model and the distance between the candidate point and the Pareto frontier, the problems that in the prior art, multi-target / multi-constraint / cross-material migration and online disturbance self-adaption are difficult to solve, and the test cost is difficult to systematically reduce under the small sample condition are solved.

The invention relates to the technical field of ultra-precision intelligent manufacturing, in particular to an ultra-precision machining error compensation method based on multi-modal information fusion, a medium and equipment. The method comprises the following steps: acquiring three-dimensional shape data of a workpiece through an in-situ measurement device, synchronously acquiring temperature field and vibration data, constructing a multi-modal space-time tensor, inputting a physical information neural network to decouple a static geometric error, a time-varying thermal drift error and a dynamic vibration error, and generating a four-dimensional correction tool path to realize reverse compensation. According to the method, physical constraints are embedded through a PINN model, the problem of false thermal drift misjudgment in a traditional method is solved, experiments show that the RMS error of free-form surface machining is reduced to 5 nm or below, and the precision is improved by 70% or above; and meanwhile, error decoupling is endowed with physical interpretability, the small sample generalization ability is enhanced, and a technical basis is provided for process optimization.

The invention belongs to the technical field of uncertainty probability characterization analysis, and discloses a thermoplastic composite material leading edge structure design allowable value probability characterization method and system under a small sample condition, and the method comprises the steps: defining a plurality of candidate probability distribution models; fitting each model based on the original sample data and calculating an AIC value and a BIC value; a dynamic weight factor alpha is calculated according to the sample size n, and then a hybrid information criterion HIC value is calculated; generating a plurality of sample sets through Bootstrap self-service sampling, recalculating the HIC value on each sample set, and counting the selected optimal frequency of each model; and determining an optimal probability distribution model according to the frequency, wherein the optimal probability distribution model is used for representing a design allowable value. According to the method, the dynamic weight factor alpha is introduced, AIC and BIC criteria are effectively unified, optimal balance between prediction precision and model complexity is achieved under the condition of small samples, and engineering practicability and robustness are remarkably improved.

The invention belongs to the technical field of steel defect detection in a real industrial scene, and particularly relates to a steel defect detection method combined with an active learning strategy, which comprises the following steps: constructing a steel surface defect data set; performing initial training based on the lightweight YOLO model; screening out a high-value sample from an unlabeled steel image through an uncertainty evaluation strategy; preferentially performing expert labeling on the selected samples to form an updated training set; performing model training on the updated training set and dynamically optimizing the network weight; a new sample is introduced into each round of iteration, an active learning closed-loop system is continuously improved, and the detection precision under the small sample condition is improved. According to the method, the manual marking cost is effectively reduced, the cross-scene adaptive capacity is improved, and a feasible path and theoretical support are provided for construction of a large-scale defect detection system in the iron and steel industry.

The invention discloses a microbe grouping analysis method and related equipment. The method comprises the following steps: acquiring microbe data of a target research area; based on microorganism abundance data, a conversion value of each sample is obtained through center logarithm ratio conversion processing, and then a distance value between any samples is obtained through Euclidean distance quantization; adaptively setting a minimum sample number and a neighborhood radius based on a sample number and a species number corresponding to the microbial data; performing density clustering on samples in the microbial data to obtain a clustering result, and further marking the sample type of each sample in combination with the minimum sample number and the neighborhood radius; performing multi-level analysis on the clustering result based on the sample type to obtain an analysis result of the clustering result; and performing correlation analysis based on the analysis result and the methane leakage data to obtain a microbe grouping analysis result of the target research area. The method realizes standardization and automation of the analysis process, can remarkably improve the analysis efficiency, and can be widely applied to the technical field of data processing.

The embodiment of the invention discloses a small sample incremental learning and training method and device based on multi-modal prompt learning, and the learning method comprises the steps: obtaining a to-be-detected original image; determining a first region-of-interest (RoI) feature of a first target object in the original image; determining a matching degree between the plurality of preset multi-modal prototypes and the first RoI feature of the first target object; based on the matching degree between each preset multi-modal prototype and the first RoI feature, performing feature fusion processing on each preset multi-modal prototype to obtain a first fusion feature; determining a second RoI feature of the first target object based on the first fusion feature and the first RoI feature; and determining a target detection result of the first target object based on the second RoI feature of the first target object. According to the method and the device, the RoI features can be optimized based on the multi-modal prototype, so that the accuracy of target object detection is improved.

The invention discloses an industrial part small sample target detection method based on synthetic data, and the method consists of a synthetic data generation module and an enhanced target detection model named YOLO-DC, and comprises the steps: separating a training process of the target detection model from dependence on large-scale real labeled data; and training the YOLO-DC model only by using virtual data generated by the synthetic data generation module. The trained model can be directly deployed in a real physical environment to accurately detect industrial parts, so that a visual perception task is completed. According to the method, a user can quickly deploy a detection system adaptive to new parts without collecting and marking real images, so that the development and deployment period of an industrial visual system is shortened, the data cost is remarkably saved, and meanwhile, the flexibility and the intelligent level of a manufacturing system are greatly improved.

The invention discloses a small sample non-uniform multi-mode clutter modeling and partition suppression method, which is applied to the technical field of radars, performs multi-mode clutter modeling aiming at a complex environment, partitions the non-uniform clutter environment based on the difference of non-uniform clutter covariance matrix geometric matching, and improves the robustness of the multi-mode clutter modeling. Then, the problem that the performance of a traditional clutter suppression method is reduced under the small sample condition is solved through a dimensionality reduction self-adaptive filtering method in different areas; according to the method, firstly, multi-mode clutter echoes are generated for a complex environment, different-mode clutter region boundaries are judged through the difference of different clutter covariance matrix geometric matching, clutters are partitioned, and then the degree of freedom is reduced through data dimension reduction in different regions, and filter weight vectors are designed to achieve self-adaptive filtering; according to the method, effective partitioning and suppression of multimode clutters can be realized under the conditions that the number of samples is less than the degree of freedom of a radar system and the clutter environment is not uniform.

The embodiment of the invention discloses a soil nutrient classification method and device, electronic equipment and a storage medium. The method comprises the following steps: acquiring soil attribute information of a target area within a preset duration; the soil attribute information comprises soil conductivity, soil temperature and spectral reflectivity; performing data preprocessing on the soil attribute information to obtain preprocessed multi-channel time sequence data; performing feature extraction and fusion on the multi-channel time sequence data based on a target feature extraction model to obtain hidden state features; the target feature extraction model is obtained by performing semi-supervised mutual learning on a preset feature extraction model by using different enhanced samples obtained by performing different data enhancement processing on unlabeled samples; and determining a soil nutrient classification result of the target region based on the target classification model and the hidden state features. Through the technical scheme of the embodiment of the invention, the soil nutrients can be accurately and conveniently analyzed, and efficient soil nutrient classification in a few-sample environment is realized.

The invention discloses a cross-domain small sample target detection method based on dynamic information fusion, relates to the field of industrial defect detection and the like, and aims to solve the problems of inter-domain distribution difference, annotation scarcity and the like in a cross-domain small sample scene. The method specifically comprises the following steps: firstly, dividing a source / target data set, constructing a DIC-ViT model containing double branches, and sharing a DINOv2 backbone network to extract features; an anchor box is generated through a region proposal network, and sample division is optimized by combining dual criteria of an IoU value and a category center distance; then feature adaptive fusion is realized through a dynamic information coupling module, and a discriminative category prototype is generated through a contrast learning module; and finally, optimizing the model by using an objective function containing classification, regression and contrast loss. The method can improve the target detection robustness and precision, is excellent in performance in a multi-data-set test, and is suitable for a detection scene with scarce samples.

The invention provides an automatic operation and maintenance method and system suitable for a closed system. The technical problem that a traditional operation and maintenance scheme is difficult to apply due to network isolation and sample scarcity in closed environments such as finance and energy is solved. The method comprises the following steps: generating a unique and traceable migration identifier for all operation and maintenance data, models and reasoning results through a data and migration management module; quantitatively calculating a transferability score between the source domain and the target domain through a transferability evaluation module; an optimal model migration strategy is dynamically selected according to the mobility score, and efficient self-adaption of the model is achieved; closed-loop operation and maintenance are executed through a multi-agent cooperation system comprising detection, diagnosis and repair agents, and self-learning and self-optimization of the system are realized through small sample active learning and a knowledge base evolution mechanism. According to the method, rapid construction, continuous evolution and whole-process traceability of the artificial intelligence operation and maintenance capability in the closed system are realized, and the method has remarkable innovativeness and industrial application value.

The method is oriented to actual scenes of rolling bearings with scarce marks in cross-equipment and cross-working conditions, and solves the problems of instable precision and negative migration caused by domain migration under small sample and semi-supervised conditions. The invention provides a cross-domain transfer learning method of domain sensing data-meta initialization-teacher and student semi-supervision-curriculum type alignment-physical consistency-online updating. According to the method, order resampling, event anchoring slicing and robust scaling are matched with a self-supervision health index and order-preserving calibration to generate a soft / interval weak label; unlabeled is absorbed through consistency learning and uncertainty gating, and domain invariant representation is obtained according to low / middle / high level step-by-step alignment; negative migration is inhibited by combining monotone / integral consistency of the non-negative degradation rate with time deformation and spectrum keeping consistency, and rapid adaptation is realized by matching meta-learning and online small-step fine tuning; the resulting generic characterization and reusable initialization can be used for health grading, phase identification, and life-related estimation.

The invention relates to a human body multi-cycle gait track prediction method based on small sample space-time decoupling. The method comprises the following steps: firstly, obtaining a single-cycle gait sample through kinematics modeling and cycle segmentation, and extracting, expanding and normalizing parameters such as stride and stride frequency; secondly, establishing a gait parameter joint distribution model by using a Gaussian mixture model and Gaussian mixture regression, and predicting gait parameters according to the target speed; and then space and time decoupling reconstruction is carried out, a high-fidelity joint trajectory is generated by adopting an adaptive collaborative attention LSTM network in space, a zoom factor is calculated based on a predicted stride frequency and a target speed in time, a time sequence is adaptively adjusted through a dynamic motion primitive, and finally a multi-period and multi-speed high-fidelity gait trajectory is generated through fusion. The method has high precision, interpretability and strong generalization ability under the small sample condition, the generated gaits are natural and coordinated and accord with the biomechanical law, and the application cost of a humanoid movement device is remarkably reduced.

The invention discloses a two-stage multi-mode bearing fault diagnosis method based on a pre-training large model, and belongs to the technical field of bearing fault diagnosis. The method aims at solving the problems that a traditional method is poor in generalization and poor in robustness under multiple working conditions and small sample conditions. The method comprises the following steps: firstly, constructing a learnable multi-modal Tokens which comprises a multi-scale patch Token, a feature Token and a fault Token, and realizing efficient extraction and fusion of multi-modal features; a time-frequency semantic fusion module is introduced, and comprehensive time-frequency features are output through adaptive frequency coding, time coding and multi-modal fusion; and inputting the multi-modal feature sequence into a pre-training BERT model, and adopting a two-stage training strategy, in the first stage, performing self-supervised pre-training by taking mask signal reconstruction as a target, and in the second stage, performing parameter fine tuning by taking fault classification as a target. According to the method, the diagnosis accuracy and the cross-working-condition generalization ability under the small sample condition can be remarkably improved.

The invention discloses a micro-seismic signal identification method based on transfer learning and BiLSTM-DCNN, belongs to the technical field of mining engineering micro-seismic monitoring and signal processing, and solves the problems of data scarcity in the initial stage of mine monitoring and low identification precision under the condition of small samples. Firstly, Mel spectrum feature extraction is carried out on a micro-seismic signal; the method comprises the following steps: constructing a BiLSTM-DCNN model comprising a bidirectional long-short term memory network and a deep convolutional neural network, and carrying out pre-training by using large-scale source mine data; and the model is adapted to target mine small sample data through transfer learning, and model parameters are finely adjusted, so that high-precision classification of the signals is realized. According to the method, the recognition accuracy and the model generalization ability of the micro-seismic signals under the small sample condition are remarkably improved, the test accuracy reaches 0.9444 and is improved by 80.85% compared with an unmigrated model, and the method is suitable for an intelligent early warning and safety monitoring system for mine dynamic disasters.

The application belongs to the technical field of biology and specifically relates to a method for analyzing glycans on single extracellular vesicles in a fluid sample. The method uses magnetic nanoparticles to capture and manipulate EVs, based on the specific affinity of lectins for specific glycan structures, uses polydisperse microdroplets to separate the captured EVs, realizes local signal amplification in the droplets through enzymatic reaction, adopts fluorescence microscopy to collect images of the droplets, and thus realizes identification and analysis of the glycan components on single EVs. The preparation process of the microdroplets formed by the oscillating emulsification method of the application makes the entire detection process more simple and fast, does not require any microfluidic chip and special device, and can be completed in a centrifugal tube, and all the microdroplets are generated in parallel and quickly, and the amount of microdroplets can be flexibly expanded. Moreover, compared with the traditional ELISA, the method has high sensitivity, low detection limit and small sample consumption.

The invention discloses an equivalence evaluation method for testing the fracture behavior of an irradiation metal material based on a small sample, belongs to the field of mechanical property testing, and deeply researches the size effect problems of non-uniform deformation and crack initiation of the metal material under the irradiation condition through three research means of experiments, theories and simulation. And a powerful basis is provided for performance evaluation of the metal material under the irradiation condition. In the aspect of experiments, micro-stretching and micro-cantilever beam experiments of samples with different sizes under irradiation conditions are carried out, and a correlation mechanism between irradiation material crack initiation and sample sizes is analyzed. On the theoretical aspect, a dislocation channel-grain boundary interaction model is established, and prediction of fracture toughness of irradiation samples of different sizes is achieved. In the aspect of simulation, a finite element simulation system for non-uniform deformation, crack initiation and size effect of the metal material under the irradiation condition is established, and the influence of irradiation and sample size on microstructure evolution and macroscopic fracture behavior of the material is explored.

The invention discloses a cross-working-condition bearing fault diagnosis method based on first-order element learning under a small sample. The method comprises the steps that original vibration signals of a bearing under different working conditions are collected and subjected to normalization processing; performing equal-length truncation sampling on the collected signals to obtain a metadata set; dividing the collected metadata set into a meta-training set, a meta-verification set and a meta-test set according to different working condition categories; constructing a fault diagnosis model based on first-order element learning Reptile; setting initial parameters of a first-order meta-learning Retile fault diagnosis model under task distribution of known working conditions; randomly sampling an N-way K-shot fault classification task, training an inner-layer parameter of a basic learning device by using the fault task, and training an outer-layer parameter of the basic learning device by using a gradient from an initialization parameter to a training weight of the fault task; and selecting the fault diagnosis model with the highest fault classification accuracy in the meta-verification stage. According to the method, the model is trained by a small number of fault samples, and the fault diagnosis of the rolling bearing is fast and accurate under the unknown working condition.