168 results about "Adaptive weighting" patented technology

The application discloses a kind of millimeter wave radar living body identification methods based on neural network, comprising: obtaining package original ADC data, signal pre-processing is carried out, and three-dimensional radar point cloud is generated;Three-dimensional radar point cloud is respectively subjected to space geometry feature enhancement and dynamics mode feature enhancement, and enhanced feature is obtained;Three-dimensional radar point cloud and enhanced feature are parallelly input into multimodal heterogeneous feature decoupling network, and the deep physical property feature of target is respectively decoupled and extracted;Cross-modal interaction and adaptive weighted fusion are carried out to the deep physical property feature, and obtain fusion feature vector;Fusion feature vector is input into classifier, and the discrimination result whether the target in package is living body is output.The deep physical property feature of target is extracted by the multimodal heterogeneous feature decoupling network of the application, and adaptive depth fusion is carried out using cross-modal attention mechanism, so as to effectively suppress non-living interference in complex dynamic environment, and the accuracy of living body detection is significantly improved.

This invention discloses a multi-condition fault detection method for inverter systems based on adaptive multi-domain unentangled manifold networks. Addressing the issue of multi-condition operation and scarce samples for some conditions in inverter systems, the method first constructs a dual-path encoder comprising a shared encoder and a private encoder to explicitly separate common features that do not change with the operating conditions from unique features that do change with the operating conditions. Then, an adaptive weighting mechanism based on reconstruction differences is used to quantify the reference value of the source domains, selectively increasing the proportion of highly correlated source domains to suppress negative migration. Next, manifold regularization techniques are used to constrain the latent space, constructing a geometrically coherent healthy manifold benchmark. Finally, the distance of the test sample projected onto the local manifold space is calculated as a fault monitoring index to achieve the fault detection objective for the inverter system.

This invention proposes a simulation method and system for the current response of distribution networks based on multi-frequency coupling effects, belonging to the field of power system modeling and simulation technology. The method includes: obtaining the frequency domain admittance matrix of the distribution network; extracting the same-frequency admittance vector and coupled admittance vector; converting each admittance vector into a transfer function, and then converting the transfer function into a discrete state-space model; preprocessing the voltage input signal using dq transform to reconstruct components at specified frequencies; substituting the preprocessed voltage signal into the discrete state-space model to obtain the current response of each frequency component; solving for the optimal weighting coefficients using an adaptive weighting strategy, and linearly superimposing the current responses to obtain the real-time simulated total current response.

This invention discloses an inversion method for S-wave velocity structure based on dense micromotion observations, belonging to the field of geophysical exploration and engineering geological exploration technology. The method includes: acquiring micromotion data; performing segmented preprocessing and cross-correlation calculations; constructing a phase consistency metric function based on the instantaneous phase information of the cross-correlation function for each time period; generating adaptive weighting coefficients; and then constructing an empirical Green's function through weighted superposition. Subsequently, time-frequency analysis is performed to obtain a time-frequency energy map; preliminary extraction of candidate dispersion velocities for each frequency is performed, and their comprehensive confidence level is calculated; after verification, a reliable dispersion curve is obtained. Based on the dispersion curve, a model parameter vector is established, and a joint objective function is constructed. A differential evolution algorithm is used for parallel global optimization inversion to obtain a one-dimensional S-wave velocity structure model below each station pair. The one-dimensional models are interpolated and fused to generate a two-dimensional S-wave velocity structure profile. This invention has the advantages of high inversion stability, high computational efficiency, and a high degree of automation.

This invention discloses an artificial intelligence-based method and system for visualizing urban planning, belonging to the field of data processing technology. The method includes: acquiring multi-source heterogeneous urban basic data and planning schemes to be evaluated; constructing a heterogeneous attribute spatial map through temporal denoising and feature matching; using the planning scheme as a priori excitation condition, inputting it along with the map into a multi-layer spatiotemporal perception graph neural network model with topology preservation loss, and outputting a spatiotemporal evolution prediction matrix tensor; constructing an adaptive weighted grid evaluation model based on multi-dimensional constraints such as perspective, predicted polarization gradient, and hardware load, and combining it with PID closed-loop monitoring of dynamic threshold, to reconstruct a high-efficiency urban visualization 3D chassis model with weighted allocation in real time. This invention effectively solves the problems of false connectivity distortion in traditional topology prediction and computational power collapse caused by rigid LOD distance pruning, achieving deep collaboration between highly realistic situational evolution and adaptive computational power.

The application discloses a self-adaptive weighted dependent code reconstruction method and belongs to the technical field of software engineering automation. The method comprises the following steps: generating an initial syntax tree by using a general syntax parser, dynamically binding an extensible attribute container to each node, performing enhanced parsing on metadata, and constructing an improved abstract syntax tree RAST; generating a composite feature vector based on a parent node influence factor and a child node influence factor; combining control flow, data flow and mixed dependent path analysis to construct a self-adaptive weighted dependent graph AWDG; introducing a semantic stability factor and a type correction coefficient, dynamically sensing the semantic stability of the node, adaptively adjusting the reconstruction recommendation strength, and calculating the reconstruction priority weight; and based on the dependent closure strategy and the reconstruction simulation pre-performance mechanism, the atomic transformation is executed after verifying no conflict in the virtual environment. Through the structural expansion and multi-dimensional weighting of the abstract syntax tree, the accurate positioning and full-process risk control of the code reconstruction are realized.

The application relates to the technical field of power system fault diagnosis, and discloses a flexible DC power distribution network fault diagnosis method and system. The method comprises the following steps: collecting a fault current time sequence signal of a DC side of a flexible DC power distribution network and inputting the fault current time sequence signal into a pre-trained fault diagnosis model; using an interpretable complex time-frequency convolution layer to perform adaptive time-frequency feature extraction on the fault current time sequence signal, obtaining a time-frequency-amplitude tensor, constructing a time-amplitude trajectory, a time-frequency trajectory and a frequency-amplitude trajectory, and respectively performing Gram difference angle field coding on the trajectories to generate a three-channel time-frequency feature matrix; inputting the three-channel time-frequency feature matrix into a parallel double-branch feature extraction module to extract global features and local features; adaptively fusing the extracted global features and local features, classifying the fused features, and outputting a diagnosis result of a fault category. The application improves the accuracy of flexible DC power distribution network fault diagnosis.

The application relates to a metamaterial phased array super-resolution imaging method based on joint regularization, which comprises the following steps: obtaining echo signals of each distance unit in a metamaterial phased array scanning radar, and establishing an imaging model between the echo signals of any distance unit and target scattering coefficients; constructing a target function containing multiple regularization terms for the imaging model, wherein the target function contains a data fidelity term for suppressing noise and a weighted L 1-norm term for adaptive group sparsity of structural characteristics of an imaging area L 2,1 norm term and a directional diagram error constraint term; iteratively solving the target function by using an alternating direction multiplier method, and outputting a super-resolution target scattering characteristic reconstruction image. The scheme fully integrates adaptive weighted L 1-norm terms and adaptive group sparsity L 2,1 norm terms and a joint correction and reconstruction mechanism, so that the scheme can cooperatively improve structural fidelity and system robustness of imaging.

This invention relates to a dynamic optimization method for apparel supply chain inventory that integrates intelligent forecasting. The invention synchronously collects multi-source data, including structured fulfillment data, customer behavior maps, and strategic configuration data. It generates a channel value tensor through normalization and semantic fusion, and introduces gating activation and dynamic attention mechanisms to achieve adaptive weighting of multi-dimensional business indicators. The replenishment optimization objective function integrates time-varying weights and channel fairness constraints, and outputs differentiated replenishment strategies through convex programming. By monitoring channel replenishment revenue deviations in real time, it applies reverse fine-tuning and learning mechanisms to continuously optimize channel value representation and replenishment decisions. This invention combines a time-aware attention mechanism with a strategy-coupled optimization framework, achieving a sensitive response to real-time business conditions and a multi-objective collaborative balance in replenishment decisions, significantly outperforming the lag and rigidity problems of traditional fixed-rule or offline optimization models.

The application provides a photovoltaic fault diagnosis method and system based on multi-modal adaptive weighting, relating to the technical field of fault diagnosis, comprising: obtaining historical operation data of photovoltaic equipment for splicing to obtain first features, clustering the first features to obtain multiple working condition clusters and corresponding order degrees; calling different feature extraction networks according to the order degrees to obtain second features; taking the Shapley values of the fault labels in the clusters to the second features as training targets to train a dynamic weight predictor; obtaining data to be analyzed, extracting first features from the data to be analyzed and finding corresponding working condition clusters, and outputting second feature weights using the corresponding dynamic weight predictor; inputting the second features of the data to be analyzed and the corresponding weights into a pre-constructed fault prediction model to obtain a fault diagnosis result. The application significantly improves the accuracy, efficiency and robustness of fault diagnosis by fusing multi-modal data, adaptive feature extraction and dynamic weight adjustment.

This invention belongs to the field of nasal alar pulse wave signal processing technology, and provides a method, apparatus, medium, and wearable device for nasal alar pulse wave signal processing. The method includes: extracting the respiratory cycle based on the respiratory signal; determining the length of the mean sliding filter window to filter out the DC component of the original nasal alar pulse wave signal, obtaining a filtered nasal alar pulse wave signal; extracting the amplitude of the AC component of the filtered nasal alar pulse wave signal; adjusting the current of the light source element corresponding to the pulse wave signal sensing element to make the amplitude of the AC component of the filtered nasal alar pulse wave signal greater than or equal to a preset amplitude threshold; further filtering the nasal alar pulse wave signal using an adaptive weighted moving average filtering algorithm; and determining whether a nasal alar pulse wave signal meeting preset quality requirements is obtained based on a comparison between the noise figure of the further filtered nasal alar pulse wave signal and a preset noise figure threshold.

This invention discloses a method and system for dynamic distribution of axle control power in urban rail vehicles based on real-time adhesion estimation. The method includes: acquiring real-time operating state data of the vehicle; calculating the dynamic vertical load of each axle during braking by combining the vehicle's longitudinal dynamics model; constructing an extended state observer and using it to estimate the wheel-rail adhesion and wheel-rail contact state parameters of each axle in real time, and calculating the theoretical maximum available adhesion coefficient accordingly; applying an asymmetric rate of change constraint to the theoretical maximum available adhesion coefficient to obtain the safe adhesion limit for constraint; constructing a multi-objective optimization function for braking force distribution, including a braking force tracking error objective and an adhesion utilization rate objective, and weighting the two objectives using adaptive weighting coefficients; and solving the multi-objective optimization function to obtain the optimal braking force for each axle. This invention can significantly improve the adhesion utilization rate and active safety performance of urban rail vehicles under complex road conditions, effectively preventing wheel slippage.

The application discloses a kind of based on deep manifold's electroencephalogram motor intention decoding method, system and device, it is related to brain-computer interface technical field.The method includes the following steps: obtaining original motor imagery electroencephalogram signal, and motor imagery electroencephalogram signal is preprocessed and feature extraction;Utilize spatial attention mechanism under the premise of keeping manifold geometric structure to the electroencephalogram channel correlation is weighted, then from time and spatial dimension respectively to the weighted covariance matrix sequence is mixed with information, simultaneously utilize symmetric positive definite residual link to retain underlying geometric information, obtain high-level geometric feature;Utilize tangent space mapping to convert high-level geometric feature to Euclidean space, and obtain the recognition result of motor imagery by classifier.The application can realize the adaptive weighting of electroencephalogram channel importance, can also retain electroencephalogram geometric feature while aggregating electroencephalogram time information, improves the robustness and accuracy of electroencephalogram decoding.

This invention provides a 2D convolutional spatiotemporally stimulated dynamic gesture recognition method based on contrastive learning enhancement, comprising: firstly, sampling and preprocessing a dynamic gesture video sequence to obtain multiple frames corresponding to the same gesture action; then, using a shared-parameter 2D convolutional neural network to extract spatial features frame by frame to obtain frame-level feature representations. Based on this, spatiotemporally stimulated enhancement of the frame-level features is performed through multi-scale temporal difference modeling, temporally adaptive weighted aggregation, and spatial attention stimulation to explicitly characterize the dynamic changes of the gesture action. The enhanced features within the same gesture sequence are constructed into a frame-level positive sample set, and a robust positive sample set center vector is calculated. Using this center vector as a positive anchor point, a contrastive learning mechanism based on hard negative sample weighting is introduced to constrain and optimize the feature space. During the training phase, the model parameters are optimized by jointly using classification loss and contrastive learning loss.

The application discloses a LIBS element quantitative analysis method based on double-branch feature fusion and electronic equipment, the method comprises the following steps: preprocessing the collected LIBS spectrum signal, including baseline correction and spectrum resampling; inputting the spectrum signal data after preprocessing into a double-branch feature extraction network, extracting local features through a CNN branch, and extracting global features through an MLP branch in parallel; performing adaptive weighted fusion on the local features and the global features through a gating fusion module to obtain fusion features; inputting the fusion features into a WMA-MLP model, the WMA-MLP model is based on MLP, integrates a multi-head self-attention mechanism and a residual module, is used for modeling the global dependency relationship between features, and outputs a final element quantitative analysis result. Through the parallelly arranged CNN and MLP branch feature extraction structures, more comprehensive spectrum feature representation can be obtained, and the accuracy and robustness of the spectrum analysis model are improved.

This invention relates to the field of textile performance testing and intelligent evaluation technology, specifically providing a home textile sleep-aiding performance evaluation system and construction method based on feature function indicators. The system includes: acquiring five-dimensional feature indicators (tactile, thermal humidity, pressure, interference, and hygiene) through standardized instrument testing, and automatically assigning weights using range normalization and entropy weighting; constructing a physically constrained Bayesian neural network, embedding prior knowledge of materials science and sleep physiology as regularization terms into the loss function, and outputting a sleep-aiding performance score and confidence interval; establishing an adaptive weighted graph convolutional network to achieve knowledge transfer and zero-sample prediction between different home textile products; and fitting the relationship between static indicators and environmental parameters through a dynamic environment adaptive mapping module to output a scenario-based sleep-aiding performance level. This invention integrates instrumental quantitative testing with multi-level neural networks, breaking away from traditional linear regression dependence and achieving rapid, objective, personalized, and scenario-adaptive sleep-aiding performance evaluation.

The application discloses a bridge full life cycle management and maintenance system based on digital twinning, which is used for the full life cycle management and maintenance of a cable-stayed bridge and comprises a data acquisition and fusion module, a digital twinning model construction and dynamic updating module and a structure health assessment and prediction module.The data acquisition and fusion module is used for acquiring monitoring data of the cable-stayed bridge in real time and fusing and processing multi-source data by using an adaptive weighting method.The digital twinning model construction and dynamic updating module is used for constructing and continuously calibrating a parameterized finite element simulation model based on bridge design parameters and fused data output by the data acquisition and fusion module, so as to serve as a digital twinning model corresponding to the physical bridge, and the digital twinning model can simulate material performance degradation.The structure health assessment and prediction module is used for assessing the current health state of bridge components by using residual characteristics of digital twinning model output and measured data based on the updated digital twinning model output by the digital twinning model construction and dynamic updating module.The application realizes intelligent management and maintenance of the full life cycle of the cable-stayed bridge.

This invention discloses a method for inverting the height of low vegetation based on an adaptive polarization decomposition algorithm in the field of remote sensing technology. The method first preprocesses the polarimetric interferometric SAR image and constructs a cross-covariance matrix. Then, it extracts the volume scattering phase and complex coherence coefficient using a polarization decomposition algorithm with an adaptive weighting factor. Finally, it extracts the even-order scattering phase and complex coherence coefficient using Freeman decomposition. After phase unwrapping, the vegetation height is calculated using the phase difference method. This invention overcomes the bottleneck of traditional RVOG models in low vegetation scenarios where the scattering mechanism coupling is difficult to separate, effectively reducing the aliasing effect of volume scattering and surface scattering, and achieving high-precision inversion of low vegetation height.

This invention relates to the field of new energy power prediction technology in power systems, specifically a wind power prediction method and system based on time-frequency coordination and adaptive decoupling. The method includes: acquiring historical wind power data; performing complexity-aware adaptive modal decomposition and reconstruction on the data to obtain sets of reconstructed sub-signals of different complexities; inputting each sub-signal set into a heterogeneous feature extraction network for parallel feature extraction to obtain heterogeneous features; adaptively weighting and fusing the heterogeneous features through a dynamic gated fusion network to generate fused features and obtain a preliminary prediction; inputting the preliminary prediction result into a frequency domain closed-loop correction network, correcting its spectrum through a learnable frequency domain filter, and converting it back to the time domain for residual compensation to obtain the final prediction. This invention achieves accurate decoupling, intelligent fusion, and in-depth correction of wind power sequences, thereby improving the accuracy of wind power prediction.

This invention relates to the field of hydrological and meteorological monitoring and information processing technology, and solves the technical problems of insufficient stability and accuracy of quantitative precipitation (QPE) in X-band phased array dual-polarization precipitation radar under strong attenuation, bright band crossing, complex terrain and rapid volume scanning scenarios. In particular, it relates to a method for quantitative precipitation estimation and error correction of X-band phased array radar. The method takes polarization-gated priority path attenuation correction as the core constraint, takes multi-parameter fusion based on weather type-rainfall intensity segmentation-adaptive weighting of cross-correlation coefficient as the main line, and converges through multiplicative (hourly scale) + additive (minute scale) hierarchical deviation correction closed loop, and finally outputs high spatiotemporal resolution precipitation products with pixel uncertainty and quality mask, which are suitable for real-time water conservancy operations at the watershed level and urban watershed.

The application discloses a kind of transmission equipment multimode real-time anomaly detection method, device, equipment and medium, the method is by obtaining the multimode monitoring signal of transmission equipment and is preprocessed, constructs synchronous time slice sequence;Respectively to each modal signal is carried out deep feature extraction, the structural coupling between each modal signal is modeled by cross-dimension time sequence coding mechanism, generates fusion state representation;State vector is obtained by aggregation to fusion state representation, and based on state vector, time series prediction is carried out to obtain basic prediction result;Instance-level uncertainty estimation is carried out and prediction bias is evaluated;According to the uncertainty estimation result, generate local correction and global correction, obtain the final prediction result by the adaptive weighted correction of gate fusion mechanism;The residual of final prediction and actual observation is calculated, and the abnormal score is generated based on residual and the abnormal determination output is carried out. Significantly improve the accuracy and real-time performance of transmission equipment under complex working conditions.

This invention discloses a medical image classification method based on a prior knowledge-guided dual-stream fusion network, belonging to the field of intelligent medical image analysis technology. The method includes: performing branch-specific preprocessing on the input raw medical image to generate semantic input images and structural input images adapted to different image branches; constructing a semantic alignment branch by extracting the global image semantic embedding of the semantic input image and calculating the classification output of the semantic alignment branch; constructing a structure-aware branch by extracting local structural features of the structural input image, constructing a category structural feature prototype matrix, and calculating the classification output of the structure-aware branch; and adaptively weighting and fusing the classification outputs of the semantic alignment branch and the structure-aware branch to obtain the final fused classification result. This invention overcomes the technical bias of existing technologies that only fuse at the decision level and cannot correct errors in the feature learning stage, significantly improving the pathological orientation and discriminative ability of model features in low-sample scenarios.

The application discloses a social robot behavior semantic consistency detection method for time sequence analysis and belongs to the field of social account detection.The core innovation of the method is to construct a "semantic indexed time sequence interaction graph", realize the deep fusion of semantic content and topological structure, introduce an offline reasoning mechanism, utilize a pre-training language model to construct a static semantic index matrix, adopt a double-flow architecture, fuse a multi-head graph attention network and a GRU on the behavior side, inhibit noise neighbor interference through an adaptive weighting mechanism, capture the space-time behavior evolution track of nodes, utilize a Transformer and a self-attention mechanism on the semantic side to extract deep text logic, mine potential robot gangs through the construction of a homogeneity association graph, extract cluster structure features, align "behavior-semantic" feature spaces through a cross-modal interaction module, and realize high-precision and low-cost identification of social network robot accounts in combination with a multi-task joint optimization strategy.

The present application provides a corn optimal seeding rate decision-making method based on multi-source field information fusion, comprising the following steps: obtaining multi-source data of the target field, respectively constructing yield prediction model data set and optimal seeding rate true value data set, and carrying out data preprocessing; constructing an end-to-end optimal seeding rate decision network, the decision network comprising a feature coding module, a double-dimension weight learning module and a gated weighted fusion module; completing feature embedding and global context modeling through the feature coding module, completing adaptive contribution modeling of two types of dimension representation through the double-dimension weight learning module, and completing multi-dimensional information fusion and adaptive weighting of multi-path candidate decision through the gated weighted fusion module, and finally outputting the optimal seeding rate prediction value; using a yield-guided step-by-step training architecture to train the decision network, obtaining target multi-source field information of the to-be-seeded plot, inputting the trained optimal seeding rate decision model, and outputting the optimal corn seeding rate of the corresponding plot.

The invention discloses a gastrointestinal symptom identification method and a sensing system, and relates to the technical field of health informatics and medical information processing, and the method comprises the steps: collecting excitation response and abdominal sound to calculate an acoustic transfer function, and employing a main formant frequency to invert a quality loading index representing a condensation water film effect; constructing an equalization gain based on the transfer function to carry out spectral shape correction on the abdominal sound time-frequency spectrum; constructing a symptom evidence vector and performing adaptive weighted scaling on the symptom evidence vector by using a quality loading index; and finally calculating the distance between the weighted vector and the symptom mode prototype library and mapping the distance into a prompt probability. According to the method, the problems of acoustic link drift and signal distortion caused by microenvironment humidity change in long-time monitoring are solved, and the accuracy and robustness of symptom identification are improved.

The application discloses a multi-view clustering method based on uniform tensor constraint non-negative embedding and spectral embedding, and belongs to the technical field of multi-view clustering, and comprises the following steps: S1, constructing a multi-view clustering model; S2, adjusting the multi-view clustering model by using an adaptive weighting coefficient; S3, adding constraint to the adjusted multi-view clustering model by using a third-order tensor, and obtaining a multi-view clustering model based on uniform tensor constraint non-negative embedding and spectral embedding; and S4, outputting a final clustering result according to the multi-view clustering model based on uniform tensor constraint non-negative embedding and spectral embedding. The application combines NMF and graph learning, and proposes a new multi-view clustering framework; the NMF and the graph learning are combined in a unified optimization framework, and consensus representation and view-specific complementary representation are successfully captured. In addition, low-rank tensor constraint effectively maintains high-order correlation and structural consistency between multiple views.

The application discloses a generator unified fatigue quantification monitoring method, system, equipment and storage medium, vibration, acoustic, electrical and temperature signal data of a generator body are acquired, high-discrimination feature extraction is performed on each signal data, key physical features reflecting the state of the generator are acquired, a generator unified fatigue cumulative quantification model is established, the total fatigue value of the generator at the current moment is characterized as the combination of a basic fatigue component, an event impact component accumulation and a self-recovery component, the key physical features are mapped into standardized fatigue increments and are accumulated in the time domain, and a continuous quantification index characterizing the overall health state of the generator is obtained, an adaptive weighted fusion algorithm based on dynamic confidence is adopted, the comprehensive confidence of each data source is dynamically calculated, and the weight of each data source in health evaluation and diagnosis is adaptively allocated in combination with the correlation with the current suspected fault mode, information islands are broken, and a generator health state capable of continuous and standardized quantification is constructed.

The application relates to the technical field of marine environment prediction, and discloses a two-branch two-stage sea surface temperature prediction method based on multi-element input. The method acquires and pre-processes multi-element data such as sea surface temperature, 2-meter temperature and atmospheric top incident solar radiation; a two-branch collaborative optimization deep learning model is constructed, and the model is trained; data of continuous days before the time to be predicted is input into the model to generate a future sea surface temperature prediction result. Among them, a short-term prediction branch extracts space-time features through ConvGRU and multi-scale convolution to predict a short-term result, and a medium and long-term prediction branch models long-range dependence through adaptive weighting and a Transformer encoder to predict a medium and long-term result; a future multi-day prediction is generated through self-recurrence rolling. Through two-branch collaboration, the application suppresses error accumulation, significantly improves the precision and stability of medium and long-term sea surface temperature prediction, and can provide efficient and accurate technical support for marine resource development.

Methods, systems, and computer-readable media for generating a personalized virtual network. The method acquires a request for a service associated with a user and their preferences. The method then identifies one or more conditions of the user and their propensities based on a first set of machine learning models using stored past information of the user. The method next identifies a second set of machine learning models and evaluates the weightage of each model based on the determined propensities of the conditions. The evaluated weights are applied to the second set of models to generate a personalized virtual network for the user.