2356 results about "Negative sample" patented technology

The invention discloses an archive knowledge base construction and retrieval method and system based on multi-modal data fusion. The method comprises the steps that heterogeneous archive data are cleaned, image features are extracted through CNN, text features are extracted through Transform, audio is converted into text and then subjected to similarity, a unified feature vector is generated, and metadata is constructed according to archive code association; creating a graph database instance, defining nodes and relationship types, importing entities and relationships, and storing feature vectors and metadata; the features are mapped to a high-dimensional shared semantic space, positive and negative sample pairs are constructed to update embedded layer parameters, self-attention is used in modalities, a shared attention mechanism is used between modalities, weights are adjusted according to archive features, and unified knowledge representation is generated; segmenting the steering quantity of the multi-modal data, storing the steering quantity into a database, and adopting hierarchical indexing and optimizing as required; related document fragments are retrieved through RAG technology vectors, answers are generated with the help of a large language model, and session feedback is provided. The file retrieval efficiency and accuracy are improved.

The invention provides an intelligent customer risk assessment system and method based on a large language model, and relates to the technical field of risk assessment, and the method comprises the steps: obtaining multi-modal data of a customer, carrying out the preprocessing, and extracting structured and unstructured features; constructing a hierarchical risk knowledge system, and realizing adaptive evolution of the knowledge system through a graph neural network and a generative model; constructing an initial negative sample library, and constructing a negative sample database in combination with a non-risk mode labeled by an expert and derivative layer analysis; optimizing the large language model by adopting a strong supervision, weak supervision and reinforcement learning cooperative training mechanism under each classification according to the customer type; mining risk features in a text by using the optimized large language model, processing multi-modal data through a multi-level attention network, and generating a positioning report including contradiction type coding, service influence dimension evaluation and risk level quantification; the accuracy, efficiency and flexibility of customer risk assessment are improved, and the risk management strategy is optimized.

The invention discloses a multi-label electrocardiogram classification method based on self-supervised pre-training and multi-modal semantic alignment, which belongs to the technical field of artificial intelligence, and comprises the following steps: realizing self-supervised pre-training of unlabeled data through a single-modal contrast enhancement network, generating global and local contrast views by adopting a multi-scale random cutting strategy, and classifying the global and local contrast views in a multi-scale random cutting mode; in combination with a teacher-student network architecture, the potential invariance features of the ECG signals are learned while negative sample dependence is avoided, the problem of annotation data scarcity is effectively relieved, and the feature robustness is improved. A multi-modal fusion mechanism based on label semantic guidance is provided, a time domain signal and a frequency domain time-frequency graph are mapped to a unified semantic space through fine-grained semantic alignment, local feature enhancement and cross-modal complementary information fusion are realized by using a cross attention mechanism, and the problem of semantic difference caused by modal heterogeneity in a traditional method is overcome. A multi-label comparison loss function based on a disease co-occurrence relation is proposed, a category discrimination boundary is dynamically optimized by modeling a label co-occurrence probability, the feature separability of a tail category is improved while the head category discrimination ability is enhanced, and the problem of sample category imbalance in a multi-label scene is remarkably relieved.

The invention relates to the field of deep learning and emotion recognition, in particular to an EEG emotion recognition system and method based on multi-scale space-time diagram convolution and comparative learning, and the method comprises the steps: carrying out the preprocessing of an EEG signal; constructing positive and negative sample pairs based on the emotion categories and the identities of the subjects; extracting hidden space representation of the electroencephalogram signal by adopting a stack auto-encoder, and dividing the electroencephalogram signal into a plurality of functional brain regions to obtain a brain source signal after dimension reduction; constructing an undirected graph structure of the brain source signal; constructing a multi-scale space-time diagram convolutional network as a feature extraction network, inputting an undirected graph structure, dynamically learning and updating an adjacent matrix, expressing features of brain source signals in the undirected graph structure, and extracting emotion related features; based on a multi-scale space-time diagram convolution and comparative learning framework, an EEG emotion recognition model is built and trained to eliminate individual differences. According to the method, the multi-dimensional information of the EEG signals is fully utilized, and the accuracy, robustness and generalization performance of emotion recognition are remarkably improved.

The invention provides a medical multi-hop text question and answer retrieval enhancement method based on heterogeneous knowledge. The method comprises the following steps: firstly, constructing a uniform heterogeneous graph structure based on a medical knowledge graph of a medical document, and establishing a semantic bridge through an entity-document mapping relationship; performing semantic decomposition on a complex medical problem input by a user by utilizing the large model, and iteratively generating a series of mutually independent atomic queries; searching a reasoning path in the entity sub-graph of the heterogeneous graph, and calculating a path score by fusing the weighted combination of the entity association text similarity, the entity matching degree and the path edge weight; training a retriever by adopting a marginal sorting loss function, and optimizing a retrieval effect through positive and negative sample comparative learning; and finally, calling a large model to convert the reasoning path with the highest score into a text, and extracting a document fragment corresponding to a path node. According to the method, the problems that an existing retrieval enhancement technology is insufficient in complex problem processing capacity, poor in reasoning interpretability and the like are effectively solved, and high-accuracy medical questions and answers are achieved.

The invention discloses a mineral resource intelligent prediction method and system based on multi-source heterogeneous data fusion and deep learning, and the method comprises the steps: collecting and preprocessing multi-source heterogeneous data, and carrying out the standardization processing to form a structured data set; multi-source heterogeneous data fusion: realizing data layer space registration and feature layer weight dynamic allocation through an attention mechanism multi-modal fusion module, and outputting a high-dimensional metallogenic feature vector; constructing a CNN-LSTM mixed deep learning model and completing initialization training, and outputting an initial mineralization probability graph; and establishing a dynamic updating engine, performing model increment training based on transfer learning, correcting the mineralization probability through positive and negative sample reinforcement learning in combination with a newly added data type, and outputting a time sequence dynamic mineralization probability graph. According to the method, mineralization probability dynamic evaluation and risk quantitative updating are realized, the prediction precision and the model updating efficiency are improved, the method is adaptive to a multi-stage exploration scene, and accurate real-time support is provided for exploration decision making.

The invention discloses a virtual fitting personalized clothing recommendation method based on artificial intelligence, and the method comprises the following steps: S1, collecting a user image, text description and behavior data, and generating a multi-modal user feature set; s2, constructing a multi-modal heterogeneous semantic map, and carrying out structural modeling and embedded representation; s3, graph embedding and feature fusion are carried out, positive and negative sample pairs are constructed, and semantic alignment training is carried out; s4, optimizing network structure parameters and training hyper-parameters by using a raccoon optimization algorithm; s5, calculating a semantic matching score, generating recommendation candidates, inputting body type parameters, and generating a multi-angle virtual fitting image; and S6, collecting user behavior feedback, updating a map edge weight and a training sample, and executing closed-loop optimization. The method has the advantages that personalized clothing recommendation and virtual fitting image generation based on the multi-modal features and the body type parameters of the user are achieved, a recommendation-fitting-feedback closed loop is constructed, and recommendation accuracy and user experience are improved.

The embodiment of the invention provides a landslide disaster negative sample optimization method based on an improved frequency ratio, and belongs to the technical field of data processing, and the method specifically comprises the steps: collecting disaster-pregnant environment data, and carrying out the preprocessing of the disaster-pregnant environment data, and obtaining raster data; defining two parameters of grading precision and grading bandwidth, introducing information entropy and a maximum frequency ratio to adaptively set parameter values of the grading precision and the grading bandwidth, and calculating a comprehensive frequency ratio; based on the statistical distribution of the comprehensive frequency ratio, taking a peak point as a negative sample screening threshold value, dividing a negative sample screening area based on the threshold value, then realizing calculation of a positive and negative sample ratio, and further determining a negative sample label under the constraint of a buffer area; according to the frequency ratio distribution of each raster data, an attribute interval with a frequency ratio greater than 1 is extracted as a significant feature, then the significant features of each raster data are classified, and an optimal feature element is determined through multi-collinearity analysis. Through the scheme of the invention, the accuracy and reliability of negative sample selection are improved.

The invention discloses a battery fault identification method based on probability labels and identification feature learning. The method is suitable for modeling and discrimination of various fault states in small sample and weak label scenes. The method comprises the following steps: firstly, acquiring key parameters such as voltage, current and temperature in an operation process of a battery system, and constructing standardized time sequence characteristic data; secondly, three types of pseudo labels are generated based on multi-source information such as alarm time difference, prediction residual error and mahalanobis distance, and a unified abnormal probability label is obtained through weighted fusion; constructing positive and negative sample pairs according to the difference between the tags, and introducing difficult samples with similar features but large tag difference to enhance the discrimination ability of the model; then constructing a twin neural network structure composed of shared parameter sub-networks, inputting positive and negative sample pairs for comparative learning, and extracting low-dimensional embedding features with clustering and distinguishability; and finally, through calculating a space distance between a new sample embedding vector and a known fault type, identification of a current fault type and evaluation of an abnormal degree are realized.

The invention discloses a pathological feature recognition and negative elimination method based on microscopic imaging. The method comprises the following steps: S1, collecting a pathological section image and digitally generating original microscopic image data; s2, preprocessing the original microscopic image; s3, constructing a pathological image recognition network fusing converter coding and a gating dynamic receptive field mechanism, and outputting pathological feature vectors; s4, performing context modeling through an attention guidance and category perception decoder, and outputting an image classification result; s5, constructing a discriminant boundary separation model based on positive and negative sample embedding, and performing negative exclusion judgment; s6, performing confidence coefficient weighted evaluation in combination with the uncertainty and the boundary distance, setting a dynamic threshold value, and screening out low-credibility samples; and S7, coding the classification result and the negative label into structured data, and sending the structured data to a diagnosis auxiliary system. According to the method, multi-scale modeling and a negative screening mechanism are fused, and intelligent recognition and credible diagnosis output of the pathological image are realized.

The invention provides a single-view image three-dimensional modeling method and system based on comparative learning, and the method comprises the steps: starting from a single input image, generating a multi-view candidate image through a pre-training two-dimensional diffusion model in combination with score distillation sampling; performing positive and negative sample division on the candidate image pairs based on texture fidelity and structure coherence; constructing a perceptual contrast learning mechanism, calculating perceptual loss between the images, and optimizing texture consistency in a three-dimensional content generation process; proposing a number-aware triple loss function, and dynamically adjusting the contribution ratio of positive and negative samples in training; then, a super-resolution reconstruction module is embedded to perform detail enhancement on the image, and the quality of a contrast gradient signal is improved; and finally, performing joint training on the perception comparison loss and a neural radiation field or volume rendering module to generate a high-quality three-dimensional grid or volume rendering result. The problems that in an existing diffusion generation method, multi-view textures are inconsistent, structure reconstruction is not coherent, and the semantic distinguishing capacity is insufficient are effectively solved.

The invention discloses a cross-modal heterogeneous data processing method and system. The method comprises the following steps: acquiring multi-source heterogeneous data including a vibration signal, a temperature signal, an oil spectral signal and the like, and forming a cross-modal alignment index sequence through event detection and time alignment processing; constructing a heterogeneous graph structure based on the aligned index sequence, and determining an edge weight according to correlation between modal fragments to obtain a cross-modal heterogeneous graph; performing representation decoupling on the heterogeneous graph, generating a cross-modal shared semantic subspace and a modal specific subspace, and reducing redundancy and noise interference through mutual information minimization constraint; establishing a prototype library, dynamically generating positive and negative sample pairs by adopting cross-modal contrast learning to enhance the discrimination ability, and obtaining an optimized cross-modal fusion result; and finally, outputting a diagnosis result including the fault type and severity, generating a diagnosis evidence track, and binding the key cross-modal fragment with the semantic centroid of the prototype library to realize interpretable traceability and self-correction of the result.

The invention discloses an emotion recognition method and system based on cross-modal alignment and matching enhancement. According to the method, firstly, feature extraction is carried out on text, audio and video modalities in a data set, and then a text and audio cross-modal emotion alignment module and a text and video cross-modal emotion alignment module are constructed respectively, so that cross-modal semantic alignment is realized. Constructing an emotion label matching module based on an alignment result, generating modal pairs with similar emotions but different labels by using a difficult negative sample mining strategy, and paying attention to cross-modal emotion consistency through a dichotomy task guide model; performing modal feature fusion on the three modals through a six-layer attention crossing mechanism, finally splicing feature vectors, inputting the spliced feature vectors into a long-sequence context fusion modeling module for deep modal fusion, and capturing cross-modal interaction information; and the fused features are sent to an emotion classification module, and a final emotion category recognition result is output.

A system for increasing the accuracy in summarization techniques is disclosed. The system generates a set of summaries for text. The system determines a label for each summary based on a set of composite metrics. The label for the summary indicates the truthfulness and faithfulness of the summary with respect to the text. The system determines that more than a threshold number of the set of composite metrics indicate that a first summary is assigned with a first label. In response, the system adds the first summary paired with the text as a positive sample to a dataset. The system determines that more than a threshold number of composite metrics indicate that a second summary is assigned with a second label. In response, the system adds the second summary paired with the text as a negative sample to the dataset. The system trains a summarization algorithm with the dataset.

The invention relates to the technical field of power grid intellectualization, in particular to a digital power grid asset classification and evolution monitoring method, system and device based on self-supervised comparative learning and a storage medium. The method comprises the following steps: acquiring network flow data in a digital power grid, extracting multi-dimensional features such as a time interval, a direction, a data packet length, a protocol type and an address port, and constructing an asset behavior feature matrix; then constructing a self-supervised contrast learning model, taking the communication behavior sequences of the same asset in different time periods as a positive sample pair, taking the communication behavior sequences of different assets as a negative sample pair, and training an asset embedding model through an InfoNCE contrast loss function; then, asset communication behaviors are mapped to a semantic embedding space, and asset classification and identification are carried out based on embedded vectors; and finally, carrying out evolution monitoring on the assets by adopting a sliding time window mechanism, generating an asset evolution trajectory sequence, and carrying out abnormity perception early warning. Autonomous learning without manual annotation, high-precision asset identification and real-time state evolution monitoring are realized.

The invention discloses a medical health cost prediction system and method based on multi-source data fusion, and relates to the technical field of medical data analysis, the system firstly fuses a plurality of heterogeneous medical information sources to generate structured data; then, multidimensional features related to the cost are extracted to construct a feature representation vector; and the cost prediction modeling module constructs a cost prediction model through graph structure modeling and semantic embedding, carries out joint optimization by combining a graph attention mechanism and semantic similarity, and outputs a prediction result by fusing a time sequence and static characteristics. A joint optimization algorithm of graph semantic comparison loss and prediction deviation loss is introduced into model training, and positive and negative sample pairs are constructed through a cost label distance. And the feedback optimization module triggers model updating when the prediction deviation exceeds a threshold value, and dynamically adjusts a model structure and parameters through a deviation index and a sample confidence factor. And the prediction interpretation module analyzes influence factors based on intermediate layer features or gradient propagation, outputs an interpretation report, and improves the transparency and credibility of the model.

The invention belongs to the technical field of image processing, and particularly relates to a wood surface defect detection method and device based on multi-view anomaly detection, and the method comprises the steps: obtaining a target image based on an original wood image; constructing a wood surface defect detection model comprising an encoder, a decoder and a defect marking module, constructing positive and negative samples based on the target image, and training the encoder by using the positive and negative samples through contrast learning; the output of the encoder is used as the input of a decoder, a loss function is constructed through a mean square error and a structural similarity index to train the decoder, and the decoder outputs a reconstructed image; obtaining a residual image between the reconstructed image and the target image through a defect marking module to obtain a fused residual image, and marking a pixel region exceeding a dynamic threshold as a defect; and detecting the preprocessed to-be-detected wood image by using the trained wood surface defect detection model to realize wood defect detection and pixel-level positioning.

The invention belongs to the technical field of agricultural machinery track classification, and particularly relates to an agricultural machinery track data classification method and device, electronic equipment and a storage medium. The method comprises the following steps: carrying out preprocessing and feature conversion operation on original agricultural machine GNSS trajectory data to obtain continuous and uniform time sequence trajectory data; the original agricultural machine GNSS trajectory data comprises a plurality of discrete trajectory points, and the category of the trajectory points comprises a field trajectory and a road trajectory; a trajectory classification model comprising a feature extraction module and a classification head is constructed, the time sequence trajectory data serve as training samples to train the trajectory classification model, the feature extraction module is trained in a comparative learning mode, and the classification head is trained through cross entropy loss; and deploying the trained trajectory classification model, and classifying agricultural machinery trajectory data by using the trained trajectory classification model. According to the invention, a contrast learning framework fused with the dynamic negative sample queue is designed, the classification accuracy and generalization ability are improved, and the method is suitable for track classification tasks in various complex scenes.

The invention discloses a multi-modal-based generative generalized zero sample learning method. The method comprises the following steps: S1, collecting and preprocessing image and text data; s2, optimizing image features by adopting self-supervised learning, and generating visual feature vectors; s3, text features are enhanced through context attention, and text feature vectors are generated; s4, constructing a cross-modal embedding space, aligning visual and text features, and constructing a positive and negative sample pair; s5, calculating the similarity of the sample pair, and optimizing cross-modal feature distribution; s6, introducing regularization constraint, modeling known category features by using the improved Gaussian mixture variational auto-encoder, and generating unknown category features; and S7, training a classification model, and performing category prediction in the optimized cross-modal embedding space. According to the method, visual and text features are fused, cross-modal alignment is optimized, the quality of unknown category features is improved, and the defects of an existing zero sample learning method in cross-modal alignment and generalization ability are overcome.

The invention provides a mathematical problem solving method and device based on a multi-modal large model and electronic equipment, and relates to the technical field of artificial intelligence. The method comprises the following steps: determining a mathematical element image of a mathematical problem; inputting the mathematical element image into an image coding model to obtain an image vector output by the image coding model; the image coding model is obtained by training based on a sample mathematical element image and a positive sample text description and a negative sample text description corresponding to the sample mathematical element image; inputting the image vector into a self-adaptive module to obtain an image conversion coding vector output by the self-adaptive module; the adaptive module is obtained based on training of a sample image vector and a sample text vector; determining question stem characters of the mathematical question, and inputting the question stem characters and the image conversion coding vector into the large language model to obtain a prediction answering process output by the large language model; the large language model is obtained based on sample question stem characters, sample image conversion coding vectors and sample answering process training, and the mathematical problem solving capability of the multi-modal large model can be improved.

The invention belongs to the technical field of target detection, and particularly relates to a high-speed rail dropper insulator integrity identification method and system based on improved YOLOv9, and the method comprises the following steps: collecting dropper insulator image data with preset definition, and generating a double-category annotation data set; performing negative sample screening on the double-category annotation data set, and dividing a training set, a verification set and a test set based on a preset proportion of 8: 1: 1; constructing an improved detection model based on a YOLOv9 framework; performing iterative training based on the training set, monitoring convergence based on the verification set, and outputting a trained improved detection model; inputting the test set into the trained improved detection model, judging whether the performance reaches the standard or not, and outputting the trained improved detection model; inputting a to-be-detected image in real time and outputting a detection result; and generating a defect alarm signal and a visual report based on the detection result. The method has the advantages of being high in high-speed rail dropper insulator integrity recognition accuracy and high in recognition speed.

The invention provides an information recommendation method and device based on collaborative sampling of a knowledge graph and an adjacent graph, and relates to the technical field of information recommendation. According to the method, the triple is constructed by obtaining the historical interaction data of the user, and the triple is fused with the project attribute graph to construct the collaborative knowledge graph; performing expression alignment through projection transformation to obtain initial embedding expression; performing multi-hop neighbor sampling through biased random walk to obtain a multi-hop neighbor set; determining a neighbor which is most similar to the embedding representation of the positive sample item as an anchor neighbor; then linear interpolation is carried out, and the embedding representation of the difficult negative sample is synthesized; on the basis of the multi-head graph attention network, training by adopting a positive sample and a synthesized difficult negative sample, and aggregating to obtain an updated embedded representation; and calculating an interaction probability based on the updated embedded representation, and generating a recommended item for the user. According to the method, the problems of user-project interaction sparsity and false negative example interference are effectively solved, and stronger semantic comprehension and generalization ability are provided for a recommendation system.

The invention discloses a reference video object segmentation method based on semantic consistency and motion perception. The method comprises the following steps: 1, constructing semantic prompt information and video frame information of a reference video object segmentation data set; 2, preprocessing the reference video object segmentation data set; 3, establishing a reference video object segmentation model based on semantic consistency and motion perception: designing a double-branch decoupling strategy for decoupling feature information at semantic and visual levels so as to extract static and motion information of text description and visual features; a hierarchical motion sensing module is designed to capture and align motion information between different frames, and analyze short-term and long-term motion information, so that the model obtains a sensing ability for a long-term motion mode; the semantic consistency module is designed to align semantic description and video features, so that the accuracy of target selection and the integrity of masks are improved, and false detection of negative samples is avoided; a perceptual dynamic fusion mechanism is designed to be used for embedding text information into a visual feature space, so that visual features can obtain text semantic information, and the cross-modal understanding ability of the model is enhanced; 4, constructing a loss function, updating model parameters, setting training parameters, and training to obtain an optimal weight; and 5, detecting the test set image based on the optimal weight to obtain a final segmentation result. According to the method, static and dynamic information is effectively decoupled, the sensing ability to an object motion mode is enhanced, and the segmentation performance is improved.

The embodiment of the invention discloses a text representation model training method and device, a storage medium and a terminal. Constructing an initial text representation model; obtaining a plurality of sample statement pairs, wherein the sample statement pairs are statement pairs consisting of sample query statements and sample comparison statements; training an initial text representation model based on the plurality of sample statement pairs; in the training process of the initial text representation model, the initial text representation model is controlled to execute at least two training tasks for the multiple sample statement pairs, and task loss values corresponding to the training tasks are calculated according to execution results of the training tasks, and adjusting parameters of the initial text representation model based on each task loss value until the initial text representation model converges to obtain a trained text representation model. A three-level positive sample, weak positive sample and negative sample comparative learning mechanism is adopted, the model is made to predict various output targets, loss values for the multiple output targets are integrated into a loss function, and the learning efficiency of the model is improved.

The invention discloses a network flow association method and system based on tetrad metric learning, and the method comprises the following steps: constructing a tetrad data set containing an adversarial disturbance sample through extracting the packet interval and packet size characteristics of network flow; a double-branch feature embedded network is combined with a channel attention and space attention module, so that the dynamic sensing ability of key features is enhanced; introducing a tetrad loss function to optimize a feature embedding space, and compulsorily confronting a rejection relationship between a disturbance sample and a negative sample; based on a dynamic mixed quantile method, a correlation threshold is adaptively selected, and the robustness is improved in combination with a multi-window voting mechanism. According to the method, the association discrimination of the network flow is realized by adopting a tetrad metric learning method, the anti-disturbance capability and the calculation efficiency are remarkably improved, the distribution characteristics of the original flow are not changed, and the method is suitable for a large-scale real-time anonymous network flow association scene.

The invention discloses a collaborative recommendation algorithm based on knowledge graph enhancement and negative sample contrast learning, which is realized through a knowledge graph perception embedding module, a volume accumulation enhancement module, a view contrast learning module and a self-adaptive contrast loss module, and belongs to the technical field of information recommendation. According to the method, high-order semantics are aggregated through knowledge graph perception embedding and a graph convolutional network, so that the problem of signal sparseness in the knowledge graph is effectively relieved; high-quality sub-views and a hard negative sampling mechanism are screened in combination with view contrast learning, noise interference is suppressed, and embedding representation robustness is improved; experiments show that the method is obviously superior to a mainstream model in Recall (at) 20 and NDCG (at) 20 indexes, and both recommendation accuracy and anti-noise capability are considered.

The invention discloses a joint multi-task table semantic analysis method based on a pre-training model, and relates to the technical field of natural language processing and databases. According to the method, SQL statements are crawled from a specified website, SQL and tables are converted into natural language texts through a large language model, columns and tables are extracted to form positive and negative samples, experimental data is converted into a Spider data set format, and natural language problems and database mode tasks are completed through a cue word template and a few-sample frame; then, an MLNaT model of a 12-layer relation perception Transform architecture is constructed, statements and column names spliced according to a specific format are input, three tasks of mask language setting, column prediction and SQL generation are set, and pre-training is carried out; and finally, evaluating on a Spider data set according to an accurate set matching rate, and taking an RAT-SQL as a baseline model. It is verified that the MLNaT model is superior to the reference model in the aspects of column prediction and SQL generation.

The invention discloses a data construction and fine tuning method for a knowledge retrieval model in the field of energy and power, and the method comprises the steps: carrying out the preprocessing of document data in the field of energy and power, and segmenting the document data into document segments suitable for the input of a retrieval model; performing question generation based on a large language model, generating a question-document pair positive sample set according to document fragments, and sampling to generate a question-document pair negative sample set; and in combination with contrast learning of the positive sample set and the negative sample set and LoRA parameter fine tuning, training a retrieval model, and performing problem retrieval by using the trained retrieval model. According to the method, high-quality problem-document pairs and challenging negative samples are generated, a comparative learning technology is combined, the retrieval model is optimized, the retrieval accuracy of the vector model in the field is remarkably improved, specific retrieval requirements in the energy power field can be deeply understood, and retrieval suggestions which are highly reliable and conform to habits in the field are generated.

The invention discloses a method for identifying small target features in a ray detection image, and relates to the field of image processing, and the method comprises the steps: carrying out the size unification, pixel standardization and normalization preprocessing of an original ray image; constructing a training sample set through image region cutting, and introducing a dynamic sampling strategy to realize positive and negative sample proportion adaptive control; enhancing the number and diversity of small target samples in a training set by using point-shaped and linear artificial defect generation strategies; constructing an image segmentation model and introducing feature jump connection to fuse shallow space and deep semantic information; combining Dice loss and Focal loss to form a composite loss function, and guiding the model to pay attention to a target region with a small area and weak gray level; and finally, a segmentation result is optimized through morphological processing and connected domain analysis, and structured target detection information is output. According to the invention, the recognition accuracy and integrity of the tiny target in the ray image can be effectively improved, and the adaptive capacity of the detection method to the change of the imaging quality is enhanced.

The invention discloses an intelligent interpretable traffic signal adaptive control method. The method comprises the following steps: training an intelligent agent through reinforcement learning according to environment state information of an intersection; a timing decision of each phase of the intersection is generated by using the intelligent agent; guiding the first large language model to generate pre-training data by the timing decision and the cue word to perform LoRA fine tuning on the second large language model, and inputting the cue word into the fine-tuned second large language model to enable the second large language model to generate a plurality of reasoning tracks to generate positive samples and negative samples; performing all-parameter fine tuning on the second large language model to obtain a traffic control signal decision model; and inputting the constructed cue word into a traffic control signal decision model to obtain each phase timing scheme of the intersection. According to the method, the defects that an intelligent traffic signal control algorithm based on deep reinforcement learning lacks interpretability and the cross-scene generalization ability is poor are overcome, and the method has important significance in improving the decision credibility and the deployment efficiency.