20 results about "Variational model" patented technology

A leakage sound signal denoising method based on a combination of optimized VMD and an improved wavelet threshold, for use in solving the problem in existing noise processing methods of low identification accuracy in processing leakage sound signals of water supply pipe networks. The present invention comprises: acquiring leakage sound signals of a real water supply pipe network, and analyzing noise components and ranges of the leakage sound signals; on the basis of a goshawk optimization algorithm, performing parameter optimization on the number K of decomposition modes and a penalty factor α of VMD to obtain optimal parameters, and using the optimal parameters to construct a variational model; using the variational model to decompose the leakage sound signals to obtain a plurality of intrinsic mode components; using a correlation coefficient method to screen the plurality of intrinsic mode components to obtain high-frequency components and low-frequency components; performing wavelet threshold denoising processing on the high-frequency components to obtain denoised high-frequency components; and reconstructing the low-frequency components and the denoised high-frequency components to obtain denoised leakage sound signals. The beneficial effects are that the signal-to-noise ratio of denoising processing is improved, and the identification accuracy is improved.

The invention discloses an escalator passenger abnormal behavior detection method and device, a medium and equipment, and relates to the technical field of abnormal behavior detection.The escalator passenger abnormal behavior detection method comprises the steps that an obtained to-be-detected video stream is divided into continuous frames, and an initial three-dimensional tensor used for representing passenger behavior information is obtained according to the continuous frames; constructing a joint optimization model based on a tensor ring low-rank model TRLRD and a tensor total variation model TTV; decomposing the initial three-dimensional tensor through a joint optimization model, determining a low-rank background and a dynamic foreground, and separating the low-rank background from the dynamic foreground to obtain a target foreground tensor; and inputting the target foreground tensor into a multi-person attitude estimation algorithm AlphaPose, performing attitude estimation on the passengers, determining an attitude detection result, inputting the attitude detection result into a trained convolutional neural network for abnormal behavior recognition, and obtaining a recognition result.

The invention discloses a moving target detection method based on low-rank full-connection tensor network decomposition and a total variation model, and belongs to the technical field of computer vision and real-time video analysis. Aiming at the problems of poor real-time performance, dynamic background misjudgment, serious noise interference and the like caused by neglect of space-time relevance, fixed rank constraint and high calculation complexity of a traditional method in a complex dynamic scene, the following technical scheme is provided: through full-connection tensor network (FCTN) decomposition, a common rank is utilized to realize accurate modeling of a dynamic background; interference from background leakage to foreground is reduced; incremental updating of time core parameters is carried out in combination with a sliding window strategy, and a historical space core is fixed to reduce calculation complexity; total variation (TV) regularization is combined to optimize a low-rank-sparse separation process, the continuity of a space gradient and a time gradient is restrained, the contour integrity of a moving target is enhanced, and noise is suppressed.

The invention relates to the technical field of automobile design, in particular to an intelligent generation type design method of an anti-collision beam, which comprises the following steps: firstly, performing experimental analysis on the anti-collision beam to obtain a section image comprising the anti-collision beam and performance response data corresponding to the section image; performing image recognition and text extraction on the section image; fusing the recognized image and the extracted text by using a multi-modal multi-layer fusion model to obtain a multi-modal design variable; training the constructed conditional variation network by using the multi-modal design variables and the performance response data corresponding to the multi-modal design variables to obtain a conditional variation model; generating design variables by using the conditional variation model, and predicting performance response data corresponding to the variables; optimal performance response data are screened out from the obtained performance response data, and then the optimal design scheme of the anti-collision beam to be designed is obtained. According to the method, the optimal design scheme is determined by using the conditional variation network, and the efficiency and the precision of optimization design are effectively improved.

The invention provides a curvature-enhanced large-displacement image variational optical flow method for motion estimation, relates to the field of image processing, and aims to describe the local structure complexity of an image by introducing an image contour curvature. On the basis of the curvature, limited self-adaptive weighted adjustment is carried out on the brightness invariant constraint and the gradient invariant constraint on the data item level of the opto-rheological model, so that the interference of unreliable matching in a complex structure region on optical flow estimation is inhibited; the robustness of optical flow estimation in illumination variation, complex texture and large displacement scenes is improved; and the numerical stability and convergence of the model in the multi-scale calculation process are ensured.

The application provides a curvature-enhanced large displacement image variational optical flow method for motion estimation, relates to the field of image processing, and aims to depict the complexity of local structures of an image by introducing the curvature of the image contour line, and to restrict the adaptive weighting adjustment of the brightness invariable constraint and the gradient invariable constraint on the basis of the curvature in the data item level of the variational model of the optical flow, so as to inhibit the interference of unreliable matching pairs on the optical flow estimation in the complex structure area, improve the robustness of the optical flow estimation under the scenes of illumination change, complex texture and large displacement, and guarantee the numerical stability and convergence of the model in the multi-scale calculation process.

The present invention belongs to the technical field of eliminating complex and strong background noise in gearboxes. It specifically discloses a variable-scale evolutionary adaptive denoising method, denoising system, and fault diagnosis system. The denoising method includes the following steps: S1, obtaining and initializing the original vibration signal; S2, establishing a variational model, solving the variational problem, and updating the modal parameters until convergence conditions are met; S3, using the updated modal parameters, calculating the kurtosis coefficient under each K and α condition to determine the center frequency and bandwidth of the bandpass filter; S4, constructing a signal model of the adaptive filter to obtain optimal filtering parameters; S5, based on the optimal filtering parameters, obtaining two sets of decomposed signals at different scales; S6, adaptively filtering the decomposed signals at different scales to obtain a characteristic signal after denoising. Using this technical solution, a composite fault signal separation algorithm based on variable-scale evolutionary adaptive decomposition is used to denoise the vibration signal and obtain a characteristic signal with better denoising effect.

The invention relates to the technical field of power load prediction, in particular to a collaborative variational multi-view clustering and high-order correlation fusion power load prediction method, which comprises the steps of constructing multi-view data, designing a collaborative variational model for joint modeling, mining high-order correlation, fusing multi-view information, carrying out dimension reduction processing and carrying out deep learning model prediction. According to the method, complementarity and consistency of multi-view data are balanced through the collaborative variational model, high-order correlation is quantized by using the covariance matrix, multi-view features are integrated, and prediction precision is improved in combination with principal component analysis and the GRU model. According to the method, the defects of a traditional method in multi-view data processing can be effectively overcome, the accuracy and generalization ability of power load prediction are remarkably improved, and the efficient load prediction requirement of a power system is met.

The invention relates to the technical field of power system relay protection and power distribution network automation, and discloses a power equipment real-time fault positioning method based on edge calculation, which comprises the following steps: constructing an edge calculation logic topological graph, presetting a positive sequence impedance module value, and monitoring a zero sequence current signal in real time. And when the sudden change of energy is detected, the edge node is switched to the dominant node, and the spectrum fingerprint seed frequency is extracted by using the VMD unconstrained variational model and is sent to the adjacent node. And the adjacent nodes forcibly decompose local signals in the same frequency band based on the frequency by using a CC-VMD constraint optimization model to obtain frequency domain aligned characteristic mode components. And each node calculates a normalized permutation entropy and exchanges data, and calculates a unit electrical distance entropy dissipation rate in combination with a positive sequence impedance module value, thereby determining a fault section. According to the method, the problem of feature alignment under asynchronous sampling is solved by using a frequency domain locking mechanism, and the recognition sensitivity and the positioning accuracy of the high-resistance grounding fault are improved through the entropy dissipation rate.

This invention relates to the field of automotive design technology, and more particularly to an intelligent generative design method for crash beams. First, experimental analysis is conducted on the crash beam to obtain cross-sectional images and corresponding performance response data. Image recognition and text extraction are performed on the cross-sectional images. A multimodal, multi-layer fusion model is used to fuse the recognized images and extracted text to obtain multimodal design variables. The constructed conditional variational network is trained using the multimodal design variables and their corresponding performance response data to obtain a conditional variational model. The conditional variational model is used to generate design variables and predict the corresponding performance response data. The optimal performance response data is selected from the obtained performance response data, thereby deriving the optimal design scheme for the crash beam to be designed. This invention utilizes a conditional variational network to determine the optimal design scheme, effectively improving the efficiency and accuracy of optimization design.

The application discloses a sequence data timespectrum radiation unification processing method and system under reference data constraints, and the method comprises the following steps: acquiring a remote sensing image sequence and generating a multi-source radiation normalization image database to form a multi-source reference sequence data pair; a reconstruction self-similarity normalization pre-training model is constructed, the reconstruction self-similarity normalization pre-training model comprises a decomposition and reconstruction module, a self-similarity weight matrix module and a variational normalization module, the decomposition and reconstruction module obtains a final timespectrum feature sequence, and the self-similarity weight matrix module obtains a timespectrum self-similarity weight matrix; the variational normalization module processes to form timespectrum radiation consistent sequence data; and corresponding timespectrum radiation consistent sequence data is obtained based on the reconstruction self-similarity normalization model. The application constructs a variational model framework of timespectrum normalization, cooperates self-similarity weight, timespectrum self-regression and similarity, realizes timespectrum radiation normalization of a timespectrum sequence cube, and solves timeserial remote sensing data which is accurate in trend and faithful in spectrum.

The present invention belongs to the field of non-destructive testing technology, and in particular relates to a full-matrix micro-ultrasonic leakage defect three-dimensional reconstruction method. The leakage defect three-dimensional reconstruction method realizes the precise positioning and three-dimensional morphological reconstruction of tiny complex defects in austenitic thin-walled stainless steel materials, and provides technical support for the detection and three-dimensional characterization of leakage micropores in austenitic thin-walled stainless steel materials. The leakage micropore three-dimensional reconstruction method includes the following steps: recording the A-Scan original signal and position information of each path point to obtain full-matrix micro-ultrasonic data; taking the minimum envelope entropy as the optimization goal, performing variational modal decomposition and reconstruction on the A-Scan original signal of each path point, and calculating the optimal solution of the variational model; performing two-dimensional mask extraction on the processed full-matrix micro-ultrasonic data to generate C-scan images of different depths; extracting two-dimensional defect edge masks at different depths; and splicing and volume rendering the two-dimensional defect edge mask results.

This invention discloses a panchromatic sharpening method for remote sensing images based on a domain-prior deep unfolded network. It addresses the shortcomings of existing deep learning-based panchromatic sharpening methods, such as insufficient interpretability, and model-based methods, which fail to adequately consider domain-specific prior knowledge. This invention models the panchromatic sharpening problem as a variational model (i.e., a panchromatic sharpening model) with spatial reconstruction and spectral modulation priors. These priors are extended into networked modules, upon which spatial information reconstruction and spectral information modulation modules are constructed. This invention solves the problems of low interpretability and insufficient consideration of domain prior knowledge in current panchromatic sharpening methods, thus improving performance. Extensive experiments on the GaoFen-2 and WorldView-2 satellite datasets demonstrate the effectiveness and superiority of the panchromatic sharpening model.

The invention discloses a method for removing a net-shaped shielding object in an image, and relates to the technical field of image processing, and the method comprises the steps: carrying out the superpixel segmentation of a to-be-processed image based on the energy functions of a color item and a structure item, carrying out the iterative exchange of pixels, obtaining a superpixel set, enabling a fusion energy function to be minimized through an image segmentation method, and obtaining a connected region set; sorting and screening according to the mean value and the variance of the connected regions to form a net-shaped shelter sample set; extracting a joint feature composed of a color histogram feature and a rotation invariant local binary pattern texture probability feature from the superpixel, and performing binary classification based on a support vector machine to obtain a mesh shelter mask; and iteratively repairing the occlusion area under smoothness constraint and edge continuity constraint by adopting a total variation model according to the mask, and outputting a repaired image. According to the method, the information of the image is fully utilized, the net-shaped shielding object can be removed without providing depth information from the outside, additional calculation is reduced, and the slender strip-shaped shielding object is repaired to be smoother through the total variation method.

This invention relates to the technical field of dynamic correction of brain deformation in neurosurgical navigation, specifically to a neurosurgical navigation method and system for adaptive correction of brain deformation. The method involves acquiring preoperative MRI data to extract initial cerebrospinal fluid (CSF) level, three-dimensional domain of the whole brain, and baseline elastic modulus, and recording steady-state intracranial pressure and baseline pulse wave amplitude. Intraoperatively, real-time acquisition of head tilt angle, mean intracranial pressure, pulse wave amplitude, CSF drainage velocity, and observed cortical surface displacement is performed. Based on this, the dynamic absolute height of the CSF level and equivalent elastic modulus are calculated, a total potential energy functional with dynamic physiological boundaries and stiffness constraints is constructed, and the three-dimensional displacement field is solved using observed cortical displacement as a forced boundary. Finally, an inverse addressing strategy is used to correct the images, and navigation is restored via DICOM flow propagation. This invention eliminates the accumulated errors in deep extrapolation caused by neglecting physiological constraints in traditional static models by fusing real-time intraoperative physiological data with a variational model of continuous medium mechanics, thus achieving adaptive correction of brain deformation.

The application discloses a power spot market equilibrium analysis method based on a conjecture variation model and belongs to the electrical engineering field.The method is based on the conjecture variation theory, an optimization problem of a generalized market operation subject is constructed, and optimal KKT conditions thereof are simultaneously established, so that the simulation operation of the power spot market game equilibrium clearing under the participation of multiple market operation subjects with multiple power generation technologies is realized.Firstly, market clearing conditions including energy supply and demand balance and a price formation mechanism are established.Secondly, a generalized optimization model of a single market operation subject is constructed with the maximum market income as the target.Thirdly, the optimal KKT conditions of all market operation subject optimization models and the market clearing conditions are simultaneously established, so that a nonlinear market equilibrium clearing model with partial differential terms is formed.Finally, a generalized conjecture variation parameter is defined and is used to replace the partial differential terms in the above model, so that the market game clearing under the participation of multiple market subjects and multiple power generation technologies can be quickly solved.

This invention provides a table-based question-answering method that combines relevant table data prompts, relating to the field of artificial intelligence technology. The invention constructs a liquid variational model to encode first information and extract feature vectors from the table; it then extracts global contextual information from the table's feature vectors to obtain contextual information output by a gated liquid network; based on this contextual information, it calculates the relevance score and uncertainty score between the table and the question text; it constructs a hybrid soft-hint strategy, using the relevance and uncertainty scores to adjust the importance of different parts of the table and constructs second information; based on this second information, it guides an answer generation model to generate the answer corresponding to the question text from the table. This invention is applicable to most mainstream generative models, thus greatly improving the versatility of practical deployment and enabling question-answering models to answer user questions more accurately.

The invention discloses a time-spectrum radiation uniformization processing method and system for sequence data under the constraint of reference data, and the method comprises the steps: obtaining a remote sensing image sequence, generating a multi-source radiation normalization image database, and forming a multi-source reference sequence data pair; a reconstruction self-similarity normalization pre-training model is constructed, the reconstruction self-similarity normalization pre-training model comprises a decomposition and reconstruction module, a self-similarity weight matrix module and a variational normalization module, the decomposition and reconstruction module obtains a final time spectrum feature sequence, and the self-similarity weight matrix module obtains a time spectrum self-similarity weight matrix; the variation normalization module processes and forms time-spectrum radiation consistent sequence data; and obtaining corresponding time spectrum radiation consistent sequence data based on the reconstructed self-similarity normalization model. According to the method, a variational model framework of time spectrum normalization is constructed, the time spectrum radiation normalization of a time spectrum sequence cube is realized in cooperation with self-similarity weight, time spectrum autoregression and similarity, and time sequence remote sensing data with accurate trend and spectrum fidelity is solved.

The invention provides a test question parameter consistency detection method and system for group knowledge diagnosis. The method comprises the steps that group, student, test question and answer result data are collected and preprocessed to obtain an interaction matrix; individual knowledge distribution is constructed, group knowledge generality characteristics are obtained, and group knowledge generality and dependency are captured based on an attention mechanism; then test question parameters are coded, and the answer correct probability is calculated based on a cognitive diagnosis model; by taking variation lower bound (ELBO) maximization as a target and knowledge distribution KL divergence minimization as a constraint, iteratively updating parameters of an encoder and the like; and finally, outputting a knowledge diagnosis result and completing DIF detection through layered likelihood ratio test. According to the group knowledge diagnosis and test question function difference (DIF) detection method based on the hierarchical variation model, the problem that knowledge dependence of students in a group is ignored in traditional detection is solved, the knowledge point prediction precision of group knowledge diagnosis is improved, objectivity and precision of test question DIF detection are achieved, knowledge state distribution KL divergence is reduced, and inference efficiency is improved.

The invention relates to the technical field of signal processing, and provides a variational mode decomposition method and device based on an orthogonal criterion, and computing equipment. The method is applied to multi-component non-stationary signal processing in a complex noise environment, and comprises the following steps: carrying out background equalization on an input signal, and extracting an empirical spectrum trend; the empirical spectrum trend indicates the overall fluctuation of the spectrum energy of the target signal; the target signal comprises a mechanical vibration signal, a biomedical signal, an electromagnetic wave signal, a sound wave signal or an artificial signal; determining initial parameters by using an empirical spectrum trend, wherein the initial parameters comprise a decomposition mode number, an initial center frequency of each mode and an initial bandwidth; constructing an orthogonal constraint variational model based on orthogonality of components of the input signal in a signal space, wherein the orthogonality is reflected between each mode and a corresponding residual error; and on the basis of an orthogonal constraint variation model, carrying out iterative updating on the mode, the center frequency and the bandwidth until an algorithm convergence condition is met, and obtaining a signal decomposition result.