12results about How to "Improve learning ability" patented technology

The application discloses a kind of based on multimodal data fusion pumped storage unit fault diagnosis method, the voiceprint signal of pumped storage unit, infrared thermal image and historical operation data are collected, improved unscented Kalman filtering algorithm is used to denoising processing voiceprint set, the voiceprint signal after denoising and historical operation data are trained using COMRes+Model, future voiceprint signal is predicted, improved Deeplabv3+Model is used to train image set, and the features of infrared thermal image are extracted;Voiceprint feature, historical operation data text feature and image feature are input into improved CentralNet model for multimodal data fusion, and the fault category is output through classification identification module, so as to complete the diagnosis of pumped storage unit fault;The method can effectively improve the accuracy and efficiency of fault diagnosis by the fusion of multimodal data and the optimization of deep learning model, and provide strong guarantee for the safe operation of pumped storage unit.

The invention discloses a conductor equipment core component cross-scale near-net shape manufacturing method, and relates to the technical field of additive manufacturing, and the conductor equipment core component cross-scale near-net shape manufacturing method comprises the following steps: S1, in a forming process, synchronously executing macro-scale and micro-scale monitoring, and associating two types of data through coordinate calibration to obtain cross-scale associated data; and S2, based on the cross-scale associated data, performing fusion analysis with real-time process and environmental parameters, and outputting a process parameter adjustment direction according to conductor exclusive analysis logic. According to the cross-scale near-net-shape manufacturing method for the core component of the conductor equipment, the overall manufacturing quality and production efficiency of the core component of the conductor equipment can be remarkably improved, the deviation of the size and the organization structure is recognized in real time in the forming process through synchronous monitoring and data fusion of the macroscale and the microscale, and the manufacturing quality of the core component of the conductor equipment is improved. And intelligent analysis is carried out based on the characteristics of the conductor material, and dynamic optimization and closed-loop regulation and control of process parameters are realized.

This invention discloses a multi-target tracking method based on Transformer and graph embedding, comprising: acquiring a video sequence; reading the video sequence frame by frame to obtain all frames; setting a counter cnt1 = 1; determining whether cnt1 is equal to the total number of frames in the video sequence; if not, inputting the cnt1-th frame and the cnt1+1-th frame of the video sequence into a pre-trained multi-target tracking model to obtain an allocation matrix between the target in the cnt1-th frame and the corresponding target in the cnt1+1-th frame, where the cnt-th frame is the previous frame and the cnt+1-th frame is the next frame; obtaining all targets associated with each target in the cnt+1-th frame based on the obtained allocation matrix between the target in the cnt1-th frame and the corresponding target in the cnt1+1-th frame; and constructing the tracking trajectory of the target in the cnt-th frame together with all the obtained targets. This invention can solve the technical problem that the global correlation learning used in existing Transformer-based multi-target tracking methods not only increases computational complexity but also suffers from excessive redundant computation.

The invention belongs to the technical field of contract auditing, and discloses a contract comparison and recognition method based on deep learning, and the method comprises the specific steps: 1, importing a recognition system; step 2, training based on a deep learning OCR model; step 3, performing OCR operation on the picture; 4, identifying and uploading the content; 5, contract output or modification is judged; and step 6, risk prompting. By automatically comparing contract texts, the system can quickly identify key information, the complexity of manual operation is reduced, the system has good adaptability, contracts of different fields and types can be processed without manually adjusting parameters, the manual intervention requirement is greatly reduced due to the high automation degree, contract auditing is more efficient, and the contract auditing efficiency is improved. According to the method, multi-language texts can be easily processed, the automatic data extraction function provides great convenience for subsequent data analysis, and compared with manual text transcription, the OCR technology has better cost benefits in large-scale document processing.

This invention discloses an imbalanced malware detection enhancement method based on the fusion of CWGAN-GP data augmentation and TEXTCNN–TRANSFORMER, comprising the following steps: S1, data acquisition and preprocessing: running an executable file in a controlled sandbox environment to dynamically capture its API call sequence, standardizing the API call sequence, and mapping it into a dense vector sequence; S2, data augmentation based on Conditional Wasserstein Generative Adversarial Network (CWGAN-GP) with gradient penalty mechanism: constructing a CWGAN-GP model conditioned on the category labels of minority malware classes, the model including a conditional generator G and a discriminator D with gradient penalty; inputting the minority malware samples obtained in step S1 and their corresponding category labels into the CWGAN-GP model for adversarial training until the model converges. The advantages of this invention are: high-quality data augmentation with semantic fidelity; comprehensive and complementary feature extraction; significant end-to-end performance improvement, especially in minority class identification; and strong model robustness and generalization ability.

The invention discloses a power transformer fault diagnosis method combining domain knowledge and a capsule network. The method comprises the steps of performing feature enhancement in combination with the domain knowledge; performing different normalization processing according to different characteristics of the original features and the knowledge features, and organizing the original features and the knowledge features into a two-dimensional matrix for convolution; and inputting a capsule network for training to obtain a trained network model, and diagnosing test data. According to the method, the domain knowledge is utilized to perform feature enhancement on the original data, and different normalization methods are adopted according to the characteristics of the original features and the created knowledge features, so that the interpretability and the diagnosis precision of the algorithm are improved; feature values are converted into a two-dimensional matrix, and the learning ability of the algorithm is improved through a hidden mode between convolutional layer mining features; aiming at the characteristic that each feature of the data of the dissolved gas in the transformer oil has specific significance, the capsule network sensitive to the vector is adopted, so that the relationship among the features can be reserved, and the method is very important for understanding and analyzing a complex mode in the data of the dissolved gas.

This invention discloses a one-step generative segmentation method, system, and medium for 3D medical images based on the average flow model, belonging to the field of medical image processing technology. The invention acquires the 3D medical image to be segmented and anatomical condition information, samples Gaussian noise as an initial latent variable, and inputs it into a MeanFlow network after temporal embedding. Anatomical conditions are input into a VeloMod module to generate scale and offset tensors, and the MeanFlow network features are modulated pixel-by-pixel. The average velocity field is calculated using the average flow identity, and target distribution features are generated through one-step mapping. A 3D segmentation mask aligned with the original image space is output by a 3D decoder. This invention achieves single-step function evaluation and inference, significantly improving segmentation speed and anatomical fidelity. It possesses advantages such as small-sample generalization, multimodal robustness, missing modality compatibility, and strong interpretability, meeting the needs of real-time clinical navigation, intraoperative planning, and high-throughput screening. It has significant application value in the field of intelligent 3D medical image segmentation.

The invention discloses a railway signal scheduling optimization method and system based on an AlphaEvove algorithm, and relates to the technical field of railway signal scheduling, and the method comprises the following steps: S1, constructing a track state diagram; s2, constructing a track state characterization model, a strategy network and a value network; s3, constructing a Monte Carlo tree, performing search by using selection, expansion, simulation and return stages in Monte Carlo tree search, and generating a preliminary scheduling strategy after search; s4, introducing a plurality of candidate scheduling sequences from the preliminary scheduling strategy as a population, generating new scheduling samples by using an evolution strategy, and screening the new scheduling samples and updating a strategy network based on path cost; and S5, selecting a scheduling scheme with least conflicts and minimum delay in the candidate samples as an optimal scheduling scheme, and outputting the optimal scheduling scheme. Compared with a traditional railway dispatching method based on rules or a single heuristic algorithm, the railway dispatching method has higher learning ability and generalization ability and can continuously adapt to different operation scenes.

The invention discloses an annular blowout preventer rubber core failure prediction method and system based on a neural network, and relates to the field of oil and gas exploitation equipment monitoring, and the method comprises the following steps: obtaining multi-dimensional working data of an annular blowout preventer rubber core, comprising a use time sequence, a displacement sequence, a hydraulic pressure sequence and a corresponding failure degree label sequence; performing dynamic space-time normalization processing to generate a normalized sample set fused with time relevance; performing supervised heterogeneous neural network model training; and inputting the working data of the rubber core of the annular blowout preventer to be detected into the optimized supervised heterogeneous neural network model, outputting a corresponding failure probability sequence and triggering an early warning mechanism. According to the method, by integrating multi-dimensional data and carrying out dynamic space-time normalization processing and multi-modal heterogeneous neural network modeling, the failure prediction precision and real-time performance are remarkably improved, and full-period management of the rubber core of the annular blowout preventer is achieved in combination with a grading early warning mechanism and threshold value self-adaptive updating.

This invention discloses a method for imputing missing values ​​in scRNA-seq data based on nonnegative matrix factorization, belonging to the technical field of deep learning and data imputation methods. This invention solves the problem of low accuracy in imputing missing values ​​in single-cell RNA sequencing data using existing methods. The main technical solution of this invention is as follows: Step 1: Filter cells based on gene data to obtain RNA sequencing data of the remaining cells after filtering; normalize, screen genes, and perform logarithmic transformation on the RNA sequencing data of the remaining cells after filtering to generate matrix X; Step 2: Decompose matrix X to obtain a feature matrix and a coefficient matrix; Step 3: Construct an input matrix for an autoencoder based on the feature matrix, use the input matrix as the input of the autoencoder, and output the imputed feature matrix through the autoencoder; Step 4: Perform decomposition, logarithmic reduction, and inverse normalization on the imputed feature matrix in sequence to obtain the imputed result. This invention can be applied to RNA sequencing data imputation.

This invention proposes a student book recommendation system and method based on an improved temporal graph neural network, comprising a student borrowing graph representation model generation module, a graph node embedding matrix generation module, a graph-level embedding vector generation module, a book borrowing relationship prediction matrix generation module, and a borrowing relationship decoding and reconstruction module. It is implemented through five processes: determining the student book borrowing graph representation model, obtaining graph-level embedding vectors, obtaining a book borrowing relationship prediction matrix based on an improved gated recurrent unit model, and obtaining the borrowing relationships after decoding and reconstructing the prediction matrix. This invention considers the changes in students' borrowing interests over time, thus influencing their borrowing behavior and preferences. Compared to the previous recommendation algorithm, this invention reduces computation time by 50% and doubles efficiency.

The invention provides a design method of a vehicle-mounted reflecting surface assisted far and near field sensing integrated system, and belongs to the field of sensing performance optimization of a sensing integrated network, the optimization problem is solved by adopting a multi-agent deep reinforcement learning technology, firstly, according to an optimization variable and an optimization target, it is determined that a base station and an RIS are used as two agents, and the optimization variable and the optimization target are optimized; then setting state spaces and action spaces of the agents, designing reward functions of the agents according to task objectives of the agents, and finally designing a multi-agent reinforcement learning algorithm model for the agents. The invention provides a multi-agent deep reinforcement learning-based design method for joint optimization of base station beam forming, receiving of a filtering matrix, phase shifting of a reconfigurable intelligent surface and a trolley moving track, maximization of system communication and sensing performance is realized according to system model parameters, and the communication performance of the system can be maximized.