43results about How to "Improve generalization" patented technology

The invention discloses a relay protection plug-in automatic simulation test method and system based on a universal backboard. The method comprises the steps that a relay protection plug-in to be tested is connected to the universal backboard; semantic information in the relay protection configuration file of the plug-in is analyzed and extracted, a corresponding test data model is dynamically constructed, and a matched operation scheme is generated; on the basis of the operation scheme, real-time simulation parameters are set, and a multi-dimensional excitation signal is applied to the plug-in through the universal backboard so as to simulate various operation conditions of the power system; and acquiring response data of the plug-in under the excitation signal through the universal backboard, and comparing the response data with an expected response to obtain a test result. According to the invention, plug-in level universal testing is realized, special debugging instruments of various models can be replaced by a single testing system, and the operation and maintenance efficiency and the spare part management level of the relay protection device are remarkably improved.

The application relates to the technical field of microbial agent, and discloses a plant root growth-promoting microbial agent screening method, which comprises the following steps: firstly, collecting rhizosphere microbial group data and corresponding crop growth phenotype data under different target crops, growth stages and soil environments; then, establishing a microbe-phenotype prediction model based on the data, obtaining two groups of key growth-promoting strains through characteristic importance analysis and microbial interaction network analysis; and finally, screening a microbial agent formula with the optimal expected growth-promoting effect according to the key strains and the prediction model. The application establishes a systematic plant root growth-promoting microbial agent screening scheme, can screen out efficient and synergistic dominant strain combinations by comprehensively considering the growth-promoting effects and interaction relationships of the strains, significantly improves the pertinence and effectiveness of the plant root growth-promoting microbial agent, and has a good application prospect in the field of microbial fertilizer development.

This invention provides a vehicle simulation speed correction method and system, relating to the field of vehicle intelligent dynamics modeling technology. The correction method includes the following steps: S1: Input and process simulation data output from the vehicle dynamics simulation model and corresponding real vehicle test data; S2: Construct a dynamic graph structure representing the interaction between state variables based on a preset physical coupling relationship of the vehicle powertrain; S3: Input the simulation data into a GCN and perform graph convolution operations under the constraints of the dynamic graph structure to extract graph embedding feature sequences representing the spatial dependencies between state variables; S4: Input the graph embedding feature sequences into a TCN and perform temporal convolution operations to learn the dynamic evolution law of state variables in the time dimension and output the correction amount of the vehicle simulation speed; S5: Output the result. Based on this, this invention solves the problem that existing correction methods have various limitations in practical applications.

The present application relates to a kind of integrated force arm structure, double rear axle suspension system and its vehicle, it includes: first force arm, the first force arm has first body, and from the first body towards one side extends to form two first arms spaced apart, each the first arm is rotatably connected with pin shaft, two the pin shaft on the first arm coaxially arranged and interval distribution;Second force arm, the second force arm has second body, and from the second body towards the side close to the first force arm extends to form two second arms spaced apart, two the second arm is arranged with two the first arm one by one, and each the second arm is rotatably connected to the pin shaft corresponding with the first arm;The first body extends to form first axle connecting arm towards away from the second body direction, the second body extends to form second axle connecting arm towards away from the first body direction.Two force arms integrated connection, simple structure, realize lightweight design.

The application belongs to the technical field of vehicle information anonymization detection, and particularly relates to a multi-type image anonymization annotation dataset construction and target coverage rate determination method. The method is based on a face and license plate image dataset with double annotation of theoretical anonymization region and anonymization features, introduces an anonymization feature extraction branch in an improved YOLOv5-L model and performs cross-modal feature fusion, combines a plurality of loss functions with theoretical anonymization region positioning loss as the core and a two-stage progressive training and difficult example mining mechanism, and realizes precise learning of the anonymization features. After normalizing the input anonymization image, the model inference obtains the theoretical anonymization region coordinates and maps them back to the original size, and through non-maximum suppression and matching of the IoU threshold, the region coverage rate is calculated to determine the positive detection, missed detection and statistical false detection rate. The application effectively overcomes the feature dependency failure and model robustness problem, and realizes high-precision, low-misjudgment anonymization detection and evaluation under various anonymization conditions.

The application discloses a tool wear prediction method for machine tool vibration time domain signal imaging processing, and relates to a predictive maintenance system, which comprises the following steps: collecting time domain signal data of machine tool vibration; converting the time domain signal into a frequency domain signal of a sinusoidal wave with different frequencies through Fourier transform; analyzing the sinusoidal wave frequency domain signal to obtain corresponding vibration amplitudes and phases, converting the sinusoidal wave frequency domain signal into vibration image points, and then converting vibration image point data at the same time into a vibration matrix to form a vibration image with complete machine tool vibration information at the same time; performing feature extraction on the vibration image through a convolutional neural network, and simultaneously performing model training on the convolutional neural network through an improved POWELL algorithm; and outputting the detection result to obtain tool wear data and perform predictive analysis.

The application discloses a kind of based on multi-source fusion and reinforcement learning's composite material additive manufacturing adaptive control method, first in printing process Synchronous acquisition space position, infrared temperature, tension and compression force and multi-source sensor data such as visible light image;Through the lightweight perception module constructed to each road data is parallelly processed and feature extraction, to force data adopts second-order difference to carry out state discrimination, to temperature data adopts multistage threshold segmentation and extracts forming area statistical feature, and reward feature is generated in combination with image recognition and position mutation detection;The multi-dimensional state features after fusion are combined with reward function, and the agent is trained using reinforcement learning algorithm to learn the optimal printing parameter decision strategy;The trained agent is used for real-time control, and the best printing speed and layer thickness instruction are dynamically output to the mechanical arm according to the current state. The application realizes accurate description and real-time adaptive control of complex multi-physical field coupling process, effectively improves printing quality and process stability.

The application belongs to the field of artificial intelligence and human-computer interaction, and discloses a motion recognition method and system based on a reconstruction adversarial domain adaptation network. The motion recognition method comprises the following steps: acquiring electromyographic signals generated by a living body when performing a motion; performing data processing on the electromyographic signals through a pre-trained reconstruction adversarial domain adaptation network to obtain a motion category corresponding to the electromyographic signals; the reconstruction adversarial domain adaptation network comprises the following parts: a feature extractor, which is used for extracting deep features of the electromyographic signals; a motion classifier, which is used for identifying the motion category according to the deep features; a domain discriminator, which is used for judging the domain attribute of the deep features; and a domain reconstructor, which is used for reconstructing the deep features into electromyographic signals. The application can solve the problem that the performance of the existing electromyographic motion recognition algorithm greatly decreases or even fails when facing unknown new users, thereby seriously affecting the practicability and adaptability of the algorithm.

The application discloses a speech separation method and device based on a multi-channel full convolution time domain network, comprising the following steps: (1) obtaining a plurality of noisy mixed multi-channel speech signals containing different sound sources, noise and reverberation, and taking corresponding pure single-channel speech signals as labels to form a training data set; (2) establishing a multi-channel full convolution time domain network, wherein the multi-channel full convolution time domain network comprises an encoder, a separator, a point multiplication module and a decoder, and the parameters of the encoder are fixed Gammatone filter coefficients; (3) inputting the training data set into the multi-channel full convolution time domain network for training; and (4) inputting a noisy mixed multi-channel speech signal to be separated into the multi-channel full convolution time domain network to obtain a pure single-channel speech signal after sound source separation. The application has better separation effect.

The invention provides an image classification model training method and an image classification method. The method comprises the following steps: acquiring an image classification model; wherein the image classification model comprises a frozen visual coding network, a field prompt parameter corresponding to at least one first image generation field and a classification network; in response to monitoring of a newly added second image generation domain, freezing the visual coding network, and adding a learnable domain prompt parameter corresponding to the second image generation domain and a classification network in the image classification model; based on the first sample image and the second sample image, performing incremental training on a domain prompt parameter and a classification network corresponding to a second image generation domain; therefore, full retraining of the backbone network is avoided, generalization and adaptability of the model to the heterogeneous generation algorithm are improved, and efficient and stable detection performance can also be achieved under the scene that the newly generated model only has a small number of samples.

This invention belongs to the technical field of multi-functional vehicles for underground coal mines, specifically a quick-connect self-locking method for the working mechanism of a series of multi-functional vehicles for underground coal mines. The device includes a switching device body, comprising a first side connected to the power vehicle body and a second side connected to the working mechanism. The upper part of the first side of the switching device body is detachably connected to the power vehicle body connection mechanism, and the lower part of the first side of the switching device body is provided with a power-side insert plate connecting plate, which is detachably connected to the power vehicle body connection mechanism; the power vehicle body connection mechanism is fixedly connected to the power vehicle body; the upper part of the second side of the switching device body is connected to the upper round steel of the working mechanism, and the lower part of the second side of the switching device body is detachably connected to the working mechanism; the device is controlled by a control system. This invention expands the vehicle's functions, improves vehicle utilization, enhances the versatility of the working mechanism of multi-functional vehicles for underground coal mines, and saves on coal mine equipment purchase costs.

The application discloses a multi-scale intestinal polyp segmentation method fusing an attention mechanism. The basic features of the method are as follows: 1. a multi-scale effective semantic fusion module is constructed to extract more abundant and effective multi-scale semantic information; 2. a new encoding-decoding deep network segmentation model is constructed to improve polyp segmentation accuracy; the method extracts sufficient context information and global information under different receptive fields, and filters out as many features useless for the segmentation task as possible, overcomes the defects that semantic information is limited and a large amount of redundancy exists in the traditional encoding-decoding structure, and has excellent segmentation and generalization performance for two-dimensional enteroscopy images with polyp regions of different shapes and different sizes.

This invention discloses a robust enhancement method for neural networks based on edge-conditional authentication training, comprising the following enhancement steps: Step 1, using a neural network model as the basic model architecture and initializing parameters; Step 2, setting batch normalization, setting batch normalization statistics based on clean input; Step 3, searching for L1 regularization strength until the neural network model training performance and neural network validation performance are approximately matched, where L1 regularization is the absolute sum of the neural network model weights; Step 4, entering the training loop to complete the robust enhancement of the neural network model. This invention improves generalization by performing a range search through L1 regularization strength, which is more conducive to obtaining a sparse and concise robust model, facilitating analysis and deployment; using a combination of L1 and L1 regularization for training facilitates a smooth transition from purely natural training to purely robust training, resulting in more stable neural network model training and faster convergence.

The invention belongs to the technical field of biomedical informatics and drug safety assessment, and provides a drug side effect frequency prediction method and device, electronic equipment and a storage medium, and the method comprises the steps: carrying out the weighted fusion of one-dimensional sequence features and two-dimensional structure features through employing learnable weight parameters; constructing a side effect similar graph through known drug-side effects, and initializing side effect nodes into semantic feature vectors by adopting preset medical term embedding; calculating an attention coefficient between the side-effect node and a neighbor node thereof by adopting a graph attention network, and updating a semantic feature vector; and training the drug weighted fusion result and the semantic feature vector to obtain a model for predicting the side effect frequency of the target drug, and the technical scheme of the invention realizes the accuracy and efficiency of drug-side effect frequency prediction.

The utility model belongs to the field of automobile electronic mechanical brake system discloses a kind of electronic mechanical brake system of driving and parking brake coupling based on gear belt, and the system includes: driving brake motor, belt wheel system, planetary gear train, parking brake motor and its transmission assembly, wherein, driving brake motor output end is fixedly equipped input gear, the output gear of belt wheel system is assembled in the sun gear input shaft of planetary gear train, gear belt is engaged between input gear and output gear, to realize deceleration and damping, planetary gear train is nested in the gear recess of output gear, driving brake torque is output via planet carrier output shaft, driving and parking brake share output shaft, compact structure, realize high transmission ratio, high torque output, meet electric vehicle high braking performance demand, gear belt has shock absorption and load sharing function, reduce noise, improve gear train durability life, double motor parallel arrangement and share output shaft design, significantly reduce axial and radial space, with high integration, high reliability and low maintenance cost advantage.

This invention relates to a wind power anomaly detection method based on dual-view comparative learning. The method includes the following steps: Step (1): Preprocessing the input data to generate a multi-scale hierarchical sequence; Step (2): Constructing an HPDB module, extracting features in parallel from both views, and fusing them with weights based on the spectrum; Step (3): Adopting a self-supervised learning paradigm, optimizing the network end-to-end by calculating a composite loss function; Step (4): In the testing phase, the model judges anomalies by calculating the weighted sum of the differences between the dual-view representations and the prediction error; Step (5): Using precision, recall, and F1 score to comprehensively evaluate the model's detection capability. This method aims to overcome the challenges faced by existing technologies in processing high-dimensional, non-stationary, and strongly coupled operating data of wind turbine generators, such as insufficient model generalization ability, insensitivity to weak anomalies, and low distinguishability between normal and abnormal patterns.

The invention discloses a U-Net-based geoelectric high-resistance organic pollution identification method, system and equipment. The method aims at solving the problems that in the prior art, identification of a high-resistance organic pollution area with weak signals and fuzzy boundaries in geoelectric data is low in efficiency and poor in precision and depends on manual interpretation. According to the technical scheme, the method mainly comprises the steps that two-dimensional geoelectric profile data of a target area are collected and preprocessed; constructing an improved U-Net segmentation network fused with an attention gate module so as to enhance the focusing capability of the model on weak pollution characteristics; training the network by adopting a progressive strategy; and finally, inputting to-be-identified data into the trained model, and automatically outputting a high-precision high-resistance organic pollution region segmentation map. The method realizes full-process automation of pollution identification, has the advantages of high identification precision, fast processing speed and strong anti-interference capability, and can be widely applied to pollution site investigation, environment monitoring and restoration engineering.

The utility model discloses a punching flow channel of a liquid cooling plate and the liquid cooling plate, which are arranged on the liquid cooling plate and comprise outer edge flow channels distributed along three edges of a rectangular plane section, and middle flow channels distributed at the middle parts of a trapezoidal plane section and the rectangular plane section; the first runners are distributed on the trapezoidal plane section and are respectively communicated with the middle runner and the outer edge runner; the second flow channel is distributed on the rectangular plane section and is communicated with the middle flow channel and the outer edge flow channel respectively; the third flow channel is distributed on the rectangular plane section, and the second flow channel communicates with the middle flow channel and the outer edge flow channel; the fourth flow channel is distributed on the rectangular plane section, and the second flow channel communicates with the middle flow channel and the outer edge flow channel; the corner runners are arranged on the edges of two bevel edges of the trapezoidal plane section; the outer edge runner is connected with a water inlet channel or a water outlet channel, and the middle runner is connected with a water inlet channel or a water outlet channel. The utility model solves the problem that the edge brazing of a liquid cooling plate is often opened and opened, and belongs to the technical field of battery cell module heat dissipation.

The invention relates to an industrial small target defect detection method and system based on ESA-YOLO, and belongs to the technical field of industrial defect detection. According to the method, structure optimization is carried out on the basis of a YOLOv8n network, and an edge feature interaction module EFIM is designed and constructed to enhance the feature expression ability and improve the perception effect of a model on a defect area; an agency attention mechanism AGAttn is introduced behind an SPPF structure in the backbone network, and feature selection and noise suppression capabilities are enhanced; and a feature enhancement alignment module FEAM is constructed, and multi-scale information fusion is optimized by combining deep and shallow features. Compared with other existing methods, the method has the advantages that the real-time detection speed is guaranteed, meanwhile, the detection effect on the defect small target is improved, the detection precision is improved by 5.6% and 4.6% on a steel surface defect detection data set and a printed circuit board defect data set respectively, and the method is suitable for defect detection application scenes in industrial scenes.

The present disclosure provides a camera position determination method and device, computer storage medium and electronic equipment, and relates to the technical field of computer video processing. The method comprises: obtaining a target scene image whose position information is to be determined, and determining an image feature vector of the target scene image based on an image encoder; inputting prompt templates of multiple scene categories into a text encoder to obtain text embedding feature vectors of target scene categories matched with the target scene image; determining a probability distribution between a target sub-vector and the image feature vector to obtain scene feature vectors corresponding to each target scene category; calculating a similarity between the scene feature vectors and the image feature vector; determining the scene feature vector with the largest similarity as a target scene feature vector, and determining position information of a camera corresponding to the target scene image according to a scene category corresponding to the target scene feature vector. This method can improve the efficiency and accuracy of camera position determination.

The invention provides a vehicle queue control method and system based on a physical information neural network, and relates to the technical field of vehicle queue control, and the method comprises the steps: obtaining the current state information and reference state information of a vehicle queue, inputting the current state information and reference state information into a centralized model prediction control module, calculating the optimal control sequence of future N steps, and outputting a basic control quantity; inputting the historical state information into a layered LSTM residual module, and outputting a residual compensation amount; adding the basic control quantity and the residual compensation quantity to obtain a final control quantity, and adjusting the acceleration; updating and dynamically adjusting the weight of the loss function of the physical information neural network, and constructing a joint loss function to synchronously optimize network parameters; the system comprises a state information acquisition module, a basic control quantity generation module, a residual compensation quantity calculation module, a control quantity fusion execution module, a state updating and collaborative optimization module and a queue control operation module. The method improves the control precision, anti-interference performance and small sample adaptability, and can effectively cope with variable speed or external interference scenes.

This application relates to an improved sample-free, incremental learning method for predicting materials microscopy images. The method includes: constructing a materials microscopy image prediction model; training the model using a set of materials microscopy images; extracting initial features from a first dataset using a feature extractor; expanding the initial features in a feature and label alignment module and determining the parameters of the linear layer while preserving the feature backbone of the feature extractor; optimizing the feature extractor and determining the parameters of the second linear layer in a feature optimization and alignment module; performing incremental alignment and saving the prototype of the new class; performing label alignment on the expanded pseudo-features based on the parameters of the second linear layer in a forgetting compensation and testing module; and using the trained materials microscopy image prediction model to predict the materials microscopy image to be predicted. This method can improve prediction accuracy.

The application discloses a thermoplastic composite mechanical property prediction method, device, medium and product, and the method comprises the following steps: obtaining component parameter sets matched with a thermoplastic composite to be simulated and a plurality of target parameter value sets matched with the thermoplastic composite to be simulated, constructing a matched target simulation representative volume element to perform a simulation experiment and obtaining a simulation experiment result. A matched target physical representative volume element is constructed according to each target parameter value set to perform a real physical experiment and obtain a physical experiment result, a plurality of training samples are obtained according to each target parameter value set, a corresponding simulation experiment result and a physical experiment result, a target loss function is set, a neural network prediction model is trained to obtain a mechanical property prediction model, and a parameter value set to be measured is input into the mechanical property prediction model to obtain a mechanical property prediction result. The technical scheme can improve the model prediction precision and generalization ability, and efficiently complete performance prediction of the thermoplastic composite in different scenes.

The application provides a kind of based on deep learning's radiation pneumonia prediction method and system.The method comprises the following steps: S1, obtains DICOM image, radiotherapy data, clinical data and immunotherapy time series data, and carries out pre-processing;S2, constructs multi-channel three-dimensional structure tensor based on DICOM image;S3, obtains multi-channel input tensor X according to planned CT volume, dose volume and three-dimensional structure tensor;obtain time series characteristic vector and clinical phenotype characteristic vector according to radiotherapy data and clinical data;S4, construct fusion prediction model, and process three-dimensional structure tensor, time series characteristic vector and clinical phenotype characteristic vector, obtain the risk probability of radiation pneumonia;S5, based on the dynamic update of input and re-prediction of updated patient data.The application realizes the deep fusion and dynamic update of multi-modal information, can provide high-precision, risk prediction with spatially interpretable, and provides decision support for clinical intervention.

This invention discloses a load time-series prediction method and system combining multi-scale slicing and masking. The method first preprocesses historical endogenous power load sequence data, historical exogenous covariate data, and exogenous covariate data to obtain preprocessed historical endogenous sequence data and multidimensional exogenous covariate data. Next, the preprocessed data is input into an endogenous multi-scale feature embedding module and an exogenous encoder, respectively, to extract multi-scale endogenous features and high-dimensional exogenous semantic features. Then, the multi-scale endogenous load features are fed into a fusion model to obtain a query vector, which, along with the high-dimensional exogenous semantic features, is input into an exogenous cross-attention module to generate the final fused features. Finally, the final fused features are passed through a feedforward network and a projection output layer to generate the power load prediction result for the future target time window. This invention solves the time alignment problem of heterogeneous exogenous power data and significantly improves the generalization and modeling capabilities for complex power grid dynamics.

The present disclosure provides a data processing method, comprising: performing enhancement processing on original data collected by an event information collection device according to at least one data enhancement transformation, to obtain at least one piece of enhanced data, wherein each piece of the enhanced data corresponds to at least one data enhancement transformation; obtaining at least one piece of initial prediction result corresponding to at least one piece of to-be-predicted data according to the at least one piece of to-be-predicted data, wherein the at least one piece of to-be-predicted data comprises the at least one piece of enhanced data, or one piece of the original data and the at least one piece of enhanced data; and obtaining a target prediction result according to the at least one piece of initial prediction result. The present disclosure also provides a data processing device, an electronic device and a computer readable medium.

This invention provides a method for predicting the lifespan of lithium batteries based on physical neural networks, comprising: constructing a coupled ordinary differential equation of lithium battery charge state and thermodynamic evolution, and establishing a continuous-time dynamic model of lithium battery charge change; constructing a PI-LSTM hybrid analysis framework based on physical information neural networks, and incorporating the coupled ordinary differential equation as a physical constraint term into the loss function of the PI-LSTM hybrid analysis framework; collecting time-series sensing data of lithium battery operation, training the PI-LSTM hybrid analysis framework, and completing the inversion identification of lithium battery physical parameters; based on the trained PI-LSTM hybrid analysis framework, outputting the state of charge and time to depletion of the lithium battery under the target operating condition, and completing the prediction of lithium battery lifespan; this invention aims to achieve accurate and interpretable prediction of lithium battery SOC and TTE in multiple devices, providing support for battery management of electrical equipment such as drones.

The invention discloses a neural network model training method and an ellipsometry method, and relates to the technical field of ellipsometry, the method comprises the following steps: obtaining a reference spectrum and a to-be-calibrated spectrum corresponding to the reference spectrum, the reference spectrum being obtained by measuring a sample through a reference ellipsometer, and the to-be-calibrated spectrum being obtained by measuring the sample through a to-be-calibrated ellipsometer; the to-be-calibrated spectrum and the simulation spectrum serve as training input, the to-be-calibrated spectrum and the simulation spectrum serve as training labels, a neural network model is trained, a loss function of the neural network model is converged, the trained neural network model is obtained, the loss function comprises a first loss item and a second loss item, and the first loss item and the second loss item correspond to each other; the first loss item is obtained based on the difference between the training output of the spectrum to be calibrated and the corresponding reference spectrum, and the second loss item is obtained based on the difference between the training output of the simulation spectrum and the simulation spectrum. The model can adjust the measurement spectrum of the ellipsometer to be calibrated so as to eliminate the measurement difference of the ellipsometers to be calibrated.

The invention provides a fine tuning method for a contract review large language model, and belongs to the technical field of artificial intelligence, and the method comprises the steps: S1, fusing multi-source laws and regulations, arbitration judgment elements and an industry contract model, and generating a first training data set for supervising fine tuning; s2, training an initial model by using the data set, and obtaining a supervised fine tuning model through adaptive parameter adjustment driven by task complexity and a chained fine tuning strategy decoupled at a task stage; s3, constructing a second training data set containing positive and negative example sample pairs based on expert knowledge; s4, training a reward model for compliance scoring based on the second training data set; and S5, taking the supervision fine tuning model as a strategy model, utilizing signals provided by the reward model, performing reinforcement learning fine tuning by adopting a strategy optimization algorithm with KL divergence constraint, and outputting an optimized contract review model in combination with clause level risk feedback and a high-risk clause penalty mechanism.

The application provides an audio signal processing method and related equipment, applied to the field of audio. The method comprises the following steps: training a diffusion model by taking a sample audio signal and sample noise information of the sample audio signal as conditions to obtain a noise reduction model; obtaining noise information in a to-be-processed audio signal; inputting the noise information and the audio signal into the noise reduction model to obtain a noise-reduced speech signal. The method has better generalization performance and better noise reduction effect; accordingly, from the perspective of hearing, the noise-reduced speech signal provided by the audio signal processing method has less noise, so that the quality and intelligibility of the speech signal are higher, and the user experience is better. In some embodiments, the noise reduction model outputs a probability distribution of the audio signal, and the probability distribution of the audio signal is sampled to obtain the noise-reduced speech signal. The time-frequency points of the noise-removed part can be restored while the time-frequency point distribution of the speech part is also recovered.