1080 results about "Estimation methods" patented technology

The invention discloses a three-dimensional attitude estimation method combining global modeling and local refinement, which comprises the following steps of: firstly, extracting a two-dimensional attitude sequence by using a human body video data set; secondly, inputting the two-dimensional attitude sequence into a structural modeling main branch, modeling a spatial topological relation and a time sequence dynamic state between joints, and outputting a global three-dimensional attitude sequence; and inputting the two-dimensional attitude sequence into a local refining branch, modeling dynamic change and detail information of a local area, and outputting a local three-dimensional attitude sequence. And finally, fusing the global three-dimensional attitude sequence and the local three-dimensional attitude sequence, generating a three-dimensional attitude sequence output, and completing three-dimensional attitude estimation. According to the method, the problem of insufficient cross-frame information transmission in a traditional method is relieved, and the accuracy and robustness of attitude estimation in a dynamic complex scene are remarkably improved.

The invention discloses a shield muck volume estimation method and system based on image processing, and particularly relates to the technical field of tunnel engineering monitoring, and the method comprises the steps: obtaining muck RGB and depth images through an image collection device, generating high-precision three-dimensional point cloud data through the combination of laser scanning and a multispectral technology, and calculating the muck volume through a slicing method. And fusing the mass flow and water content data of the belt weigher, dynamically generating a calibration factor, and correcting a volume estimation result in real time. The system comprises an image acquisition module, a point cloud processing module, a volume calculation module and a dynamic calibration module, has the characteristics of automation, high precision, real-time feedback and the like, effectively solves the problems of large error and slow response of a traditional manual metering and single weighing mode, and is suitable for continuous and accurate monitoring of the volume of muck in the shield construction process.

The invention discloses a generator state estimation method and system considering noise and parameter uncertainty constraints. The method comprises the following steps: acquiring model parameters and dynamic state vectors of a generator, and establishing augmented state vectors; performing unscented transformation on the augmented state vector to obtain a particle set; improving to obtain robust mixed Kalman particle filtering, and in the process of performing unscented Kalman filtering on an augmented state vector, taking correlation entropy maximization of a measurement information sequence as a target function, and solving by adopting a fixed point iteration method to obtain a filtering gain; determining a filtering gain according to the updated state of the measurement information; robust mixed Kalman particle filtering is executed, physical constraints of model parameters serve as a feasible region, after resampling, projections of the model parameters in new-generation particles exceed the feasible region, the model parameters are set to be closest boundary points, and then resampling is conducted again; a weighted average value of the particle set is an optimal joint estimation value, a dynamic state estimation value and a model parameter identification result are separated, and reliable uncertainty quantization is provided for state and parameter estimation.

The invention discloses an SVG valve hall cooling efficiency evaluation method and system based on a probabilistic graph model, and the method comprises the steps: obtaining original time sequence data which is obtained through the collection of a cooling system multi-parameter monitoring sensor group disposed in an SVG valve hall in continuous T sampling periods; preprocessing the original time series data to obtain a credible time series data set; constructing a Bayesian network topological structure comprising three-level nodes of an environment layer, a component layer and an efficiency layer and causal dependence edges, and optimizing parameters of the Bayesian network topological structure by adopting a maximum likelihood estimation method to form a dynamic Bayesian network model after parameter calibration; and the credible time sequence data set is used as an evidence variable to be input into the Bayesian network model after parameter calibration, calculation is carried out through a belief propagation reasoning algorithm, a final control instruction set is generated through probability weighted scoring processing, the final control instruction set is fed back to a valve group monitoring system, and early warning and automatic load reduction are achieved. The problems of large evaluation deviation and early warning lag in the prior art are solved.

The invention relates to the technical field of underwater positioning, in particular to an AUV (Autonomous Underwater Vehicle) three-dimensional pose joint estimation method and system based on multi-modal layering, and the method comprises the steps: distributing a historical observation sequence to corresponding independent convolutional neural network branches according to modals, carrying out the local time sequence feature extraction through each branch, and carrying out the local time sequence feature extraction; outputting local time sequence characteristics of each mode; performing hierarchical feature fusion based on local time sequence features of each mode, inputting global fusion features into a double-branch regression head, respectively decoding through a position regression branch and an attitude regression branch, and outputting a three-dimensional position increment and an attitude increment; and constructing a loss function by using the three-dimensional position increment and the attitude increment, training to obtain an optimal model, and deploying the optimal model to an AUV platform, thereby breaking through the limitation that the traditional method only focuses on a two-dimensional plane or separately estimates the attitude, completely covering the full-space positioning demand of the AUV three-dimensional maneuvering task, and improving the integrity and consistency of the attitude estimation.

The invention provides a front and rear axle load estimation method and system and a steer-by-wire vehicle, and the method comprises the steps: (1) when the vehicle is in a longitudinal stable and straight driving state, updating the mass of the whole vehicle; (2) establishing a least square recursion equation with a forgetting factor based on the dynamic motion equation of the vehicle on the ramp, and calculating the gradient of the current ramp; and (3) calculating a front axle load and a rear axle load based on the whole vehicle mass and the current gradient. The whole vehicle mass estimation is started under the working conditions of longitudinal stability and straight driving, so that the reliability and the stability of the whole vehicle mass calculation process are ensured; besides, vehicle dynamics and kinematics models are fused, recursive calculation is performed by applying a least square method with a forgetting factor, the road gradient can be estimated in real time at high frequency and high precision under various driving working conditions of the vehicle, and the limitation of a traditional fixed parameter model is broken through.

The invention discloses a sparse regularization direction of arrival estimation method based on a risk minimization principle, and belongs to the technical field of array signal processing and underwater acoustic signal processing. The method comprises the steps of receiving array signals and establishing an observation model; constructing a sparse representation and over-complete dictionary; establishing and initializing a regularization optimization model; carrying out adaptive weight updating and risk-driven parameter selection; after regularization parameters are determined, a fast iterative shrinkage threshold algorithm FISTA is adopted to carry out optimization solution, and dictionary refinement is carried out on the detected direction after each iteration convergence so as to reduce off-grid errors; and after a small amount of outer layer iteration is repeated, outputting a final DOA estimation result and corresponding power. According to the method, self-adaptive selection of regularization parameters and noise levels can be realized, and the problem of precision degradation under complex conditions of low signal-to-noise ratio, limited snapshot number, signal source correlation, power imbalance and the like is effectively solved without manual parameter adjustment, so that the robustness and practicability of estimation are remarkably improved.

The invention belongs to the crossing field of information and ocean science and technology, and discloses a DOA estimation method for a UUV broadside parallel co-prime array. According to the method, firstly, a virtual extension array is generated by utilizing an array auto-covariance and cross-covariance matrix, then, Toeplitz completion is carried out on'holes' in the virtual array through trace norm regularization and nuclear norm regularization constraints, a complete covariance structure is recovered, and finally, an extension matrix is constructed, and rotation invariance of a signal subspace of the extension matrix is utilized, so that the covariance structure of the virtual array is obtained. The pitch angle and the azimuth angle of the target are jointly solved, automatic angle pairing is achieved, and the effectiveness and the reliability of the method are verified through simulation results. According to the method, aiming at the application challenges that the UUV broadside array space is limited and the underwater environment is complex, all virtual array elements of the broadside parallel co-prime array are fully utilized, the array freedom degree and estimation precision are remarkably improved, the target detection capacity of the UUV is effectively enhanced, and the method has high practical engineering application value.

The invention discloses a transformer area topological structure estimation method, system, equipment and medium based on multi-dimensional power utilization characteristics, and belongs to the technical field of power distribution transformer area topologies, and the method comprises the steps: collecting multi-source data, carrying out dynamic weight distribution and abnormal value correction, carrying out the characteristic extraction according to the collected multi-source data, and generating a high-dimensional characteristic vector; clustering user nodes, dividing cluster labels, performing topological modeling according to a cluster division result, generating a topological graph, and performing anomaly verification through multi-dimensional anomaly scoring and abnormal power utilization detection; and performing incremental model parameter correction and multi-objective optimization according to anomaly verification feedback, and generating dynamic topological graph rendering and multi-dimensional decision suggestions by integrating topological modeling, anomaly verification and optimization results. According to the method, accurate estimation of a topological structure is realized through dynamic clustering and hidden node recognition, real-time diagnosis of abnormal nodes and self-correction of a model are realized by means of a multi-dimensional abnormal scoring and self-adaptive optimization mechanism, and the accuracy and operation and maintenance efficiency of transformer area management are improved.

The invention provides a ship distance estimation method and device based on an AIS (Automatic Identification System) supervision signal, relates to the technical field of intelligent shipping and water traffic supervision, and solves the problems that the prior art depends on a single sensor or a complex multi-source fusion scheme, so that supervision blind areas, inaccurate distance measurement and system complexity are caused, and the cost is low. And effective distance estimation cannot be carried out on the AIS-free ship. The method comprises the following steps: acquiring and preprocessing AIS data and video data; obtaining an approximate predicted position of the ship at the current moment through the AIS trajectory prediction model, and obtaining a ship target detection frame of the current field angle through the target detection model; dividing the current field angle of the camera into equal credible areas and carrying out space-time matching; and the matched visual features and AIS distance information are fused, an AIS distance is used as a supervision signal to construct a distance regression module, and a target detection model is trained to output a ship distance estimation value. The method is used in the other ship distance estimation process of the shore-based and ship bridge view angle.

The invention relates to an object attitude estimation method based on scene-level semantic three-dimensional Gaussian spatter, which comprises the following steps of: firstly, constructing scene-level three-dimensional representation according to a multi-view image, and associating semantic embedding of three-dimensional Gaussian points of each target object; positioning a target mask area of a target object described by a language instruction according to the query image, and extracting each target three-dimensional Gaussian point subset through semantic matching and three-dimensional space clustering; secondly, an ICP registration method is guided through two-stage learning, and the initial 6D pose of the target object is obtained; and finally, performing cascade fine optimization by combining pose rendering and similarity comparison under disturbance to obtain a 6D pose of the target object in the target scene. According to the design scheme, target retrieval and instance extraction under open vocabularies are achieved through three-dimensional Gaussian reconstruction of semantic enhancement, and the attitude estimation precision under the conditions of shielding and low texture in a complex scene is effectively improved by combining language prompt and registration and rendering fine optimization guided by two-stage learning.

The invention discloses a 6D pose estimation method and device fusing an attention mechanism, equipment and a medium, and relates to the technical field of object space poses, and the method comprises the steps: processing an RGB image and a depth image of a target object through a preset multi-modal feature fusion model, the model comprises a semantic segmentation module, a feature extraction module, a feature fusion module, an attitude estimation module and an attitude iterative optimization module. A target mask point cloud is obtained through semantic segmentation, feature fusion is performed by using a cross attention mechanism and deformable convolution, and a final pose is obtained through pose estimation and iterative optimization, so that the problems of insufficient feature extraction and weak multi-modal feature association in complex scenes such as weak texture and shielding are solved, and the accuracy and robustness of pose estimation are improved.

The invention discloses an unmanned aerial vehicle state estimation method, a readable storage medium and a navigation device, and belongs to the technical field of unmanned aerial vehicle navigation. The method comprises the following steps: establishing a linear state space model with multi-cluster measurement noise, modeling the measurement noise as multivariate Gaussian distribution, and introducing a measurement noise covariance matrix coefficient; joint prior updating is carried out based on posterior information of a previous moment, and prior estimation of a state, a state covariance, a measurement noise covariance matrix coefficient and a generalized inverse Gaussian (GIG) distribution parameter is obtained; performing pre-clustering on the measured values; online adaptive clustering is realized by using an EM algorithm, and unknown outdoor scene noise can be dynamically identified and divided; joint posterior updating is carried out, fixed point iteration is not needed, and therefore an updated measurement noise covariance matrix is obtained; updating the state and outputting the unmanned aerial vehicle state. According to the invention, through precise noise modeling, online clustering and adaptive adjustment of the measurement noise covariance matrix, the positioning precision and system robustness of the unmanned aerial vehicle are significantly improved.

The invention provides a high-resolution DOA (direction of arrival) estimation method and system, and the method comprises the steps: updating a model covariance matrix based on a to-be-corrected space power spectrum and noise information of the iteration of this round, thereby obtaining a direction dependence weight factor, and carrying out the estimation of the direction dependence weight factor. And according to the direction dependence weight factor of the current round of iteration, the to-be-corrected spatial power spectrum, the model covariance matrix and the sample covariance matrix, predicting the to-be-corrected spatial power spectrum and noise information of the next round of iteration, and finally obtaining the spatial power spectrum. According to the method, on the premise that the number of prior information sources is not needed, the problem of spectrum peak merging of a traditional sparse method under the condition of adjacent information sources is solved, stable separation of near-angle incoming waves is achieved, and the high-resolution performance of DOA estimation is remarkably improved.

The invention discloses a permanent magnet synchronous motor sensorless control method based on a dynamic position error. A transition stage is provided, per-unit estimation position errors between a first position error signal obtained based on a high-frequency signal injection method and a second position error signal obtained based on a sliding mode observer are compared, an optimal switching speed point is dynamically determined, smooth fusion is carried out on the two position errors by using a nonlinear weighting function, and the optimal switching speed point is obtained. And finally, the estimated position and rotating speed are obtained through a phase-locked loop. And finally, position tracking is carried out by adopting different error signals according to the estimated rotating speed. According to the method, the problem of estimation value pulsation and jump caused by the fact that a traditional method depends on fixed experience switching points is solved, smooth and self-adaptive switching of the two estimation methods in the full-speed domain range is achieved, the control precision and operation stability of the system are remarkably improved, the structure is simple, and engineering implementation is easy.

The invention discloses a power distribution network state estimation method and device based on high-order volume Kalman filtering, and the method comprises the steps: firstly calculating a volume point of a state variable of a power distribution network at a current moment, calculating a propagation volume point according to a state equation of the power distribution network, and calculating the state variable of the power distribution network according to the propagation volume point; and calculating a state predicted value at the next moment and a predicted value of the state error covariance matrix, and performing volume transformation on the propagation volume point to obtain a volume point at the next moment. Then, propagation is carried out on a volume point at the next moment through a measurement function, and a measurement predicted value at the next moment, a measurement error covariance matrix and a cross covariance matrix are calculated; and finally, calculating a Kalman filtering gain according to the measurement error covariance matrix and the cross covariance matrix, correcting a state prediction value by adopting the calculated Kalman filtering gain, and updating the state error covariance matrix, so that the state estimation precision is improved, and the estimation accuracy is improved. And the robustness and the precision of the system facing bad data are effectively enhanced.

The invention discloses a specific sea area water quality change trend prediction method based on time sequence analysis, and relates to the technical field of marine environment monitoring. The method comprises the following steps: firstly, collecting historical water quality data and time-space attributes of a target sea area monitoring station, and constructing a multi-graph structure based on geographical proximity, water quality change similarity and hydrological connectivity to represent a complex spatial dependency relationship; secondly, extracting multi-scale spatio-temporal features, and generating node embedding representation; and then, multi-graph feature aggregation is carried out by using a multi-graph convolutional network and an adaptive weight mechanism. And inputting the fusion features into a deep twin coding-decoding model to carry out multi-step water quality prediction, and finally, generating multiple groups of future scenes by disturbing key external driving factors by adopting a multivariate state estimation method, and outputting a multi-scene prediction sequence under different environmental conditions. According to the method, the accuracy, interpretability and decision support capability of water quality prediction are effectively improved.

The invention discloses a ground and satellite observation constraint combined GPP multi-task learning estimation method, system, medium and equipment, and belongs to the technical field of remote sensing, the method comprises the steps of obtaining OCO-2 SIF data and TROPOMI SIF data, outputting a global GPP product prediction value through an expert network, and obtaining a global GPP product prediction result; the processing flow comprises the following steps: extracting initial features based on an OCO-2 SIF data set and a TROPOMI SIF data set, inputting the initial features into a shared encoder module, learning cross-sensor spatial and temporal features through a feature alignment and fusion mechanism, outputting OCO-2 SIF time features and TROPOMI SIF spatial features, carrying out independent decoding and outputting predicted values, and outputting the predicted values. According to the global GPP high-precision estimation method, the problem of an error mode of high-value underestimation and low-value overestimation caused by insufficient space-time generalization due to single task model estimation is solved.

The invention relates to an all-angle hemispherical gyroscope delay and damping asymmetric error joint compensation method, which belongs to the technical field of inertial navigation, and comprises the following steps: S1, obtaining a self-precession excitation signal and an angular velocity measurement value signal output by a gyroscope; s2, preliminarily estimating the time delay of the gyroscope by adopting a maximum likelihood weighted cross-correlation time delay estimation method; s3, estimating the damping asymmetry error of the gyroscope through a recursive least square method by utilizing the estimated time delay; s4, re-estimating the time delay of the gyroscope by using the estimated damping asymmetric error to obtain the updated estimated time delay; judging whether the updated estimated delay is converged or not; if not, returning to the step S3, and re-estimating the damping asymmetric error of the gyroscope through the recursive least square method by utilizing the updated estimated delay; if the time delay is converged, the finally estimated time delay is obtained; and S5, compensating the finally estimated delay and damping asymmetric error into a control system of the gyroscope. According to the invention, the measurement precision and long-term stability of the gyroscope are improved.

The invention belongs to the technical field of crossing of information and ocean science and technology, and discloses a UUV broadside parallel array covariance weighted fusion DOA estimation method. The method comprises the following steps: firstly, calculating auto-covariance and cross-covariance matrixes of two sub-arrays, carrying out weighted fusion on covariance matrixes of different sources through differential common-array expansion of a virtual aperture, then reconstructing a complete covariance matrix by respectively adopting a convex optimization method of trace regularization and nuclear norm regularization according to the characteristics of the fused matrix, and finally, carrying out weighted fusion on the covariance matrixes of different sources. The pitch angle and the azimuth angle of the target are jointly estimated by performing subspace decomposition on the constructed extension matrix, and the effectiveness of the method is verified by a simulation result. According to the method, the precision of parameter estimation is improved through data fusion, the degree of freedom is remarkably improved under a small number of physical array elements, high-precision azimuth estimation of the UUV on multiple targets in the underwater environment is achieved, the target detection capacity of the UUV is enhanced, and the method has high engineering application value.

The invention discloses a radar short-time heavy rainfall estimation method based on classification echoes and environmental physical constraints, and belongs to the technical field of meteorological detection, and the method comprises the following steps: S1, multi-source data integrated fusion and cooperative gridding preprocessing; s2, rainfall type dynamic identification and Z-R relation self-adaptive primary selection based on multi-feature fusion; s3, adaptive correction of the estimation result driven by the environmental physical process is carried out; s4, estimating sequence optimization and systematic deviation correction based on a sliding time window and live feedback; and S5, multi-source information optimal fusion and refined heavy rainfall product generation. According to the method, the problems that a traditional fixed Z-R relation and single data source estimation method is insufficient in precision and insufficient in physical mechanism consideration are effectively solved, the accuracy of short-time heavy rainfall estimation is remarkably improved, and the method has obvious service application value.

The invention provides a vertical velocity quantitative estimation method and system based on a dual-polarization radar, and the method employs the multi-polarization parameters of the dual-polarization radar, combines a deep learning technology with radar meteorological physical knowledge, builds a nonlinear relation between the multi-polarization parameters of the dual-polarization radar and the vertical velocity, and achieves the quantitative estimation of the vertical velocity. Quantitative estimation of the three-dimensional vertical velocity profile is realized, and a quantitative analysis product is provided for convective weather forecast and early warning. According to the method, a differential reflectivity column, a differential phase column recognition algorithm and a dual-polarization radar hydrogel classification algorithm are combined to screen an upflow area of convective weather, and a data set of dual-polarization radar observation characteristic data and vertical velocity of the upflow area is established. Transform is used as a prediction model, a space-time attention mechanism is added, and multi-time step information is used as input, so that extraction of time features observed by the dual-polarization radar is facilitated, and the vertical velocity prediction accuracy is improved.

The invention is suitable for the technical field of computer vision, and provides a self-supervised monocular depth estimation method based on wavelet feature enhancement, and the method comprises the steps: firstly obtaining a two-dimensional target image and an adjacent image frame, and then generating a multi-scale first feature map and a multi-scale second feature map based on a main encoder and a wavelet feature extractor, the method comprises the following steps of: constructing a multi-scale second feature map, constructing a wavelet guide feature fusion module, injecting and enhancing high-frequency detail information of the multi-scale second feature map to a multi-scale first feature map based on the wavelet guide feature fusion module, generating a plurality of third feature maps, generating a multi-scale depth estimation map based on a decoder, and finally obtaining a multi-scale depth estimation map based on relative pose information, the multi-scale depth estimation map and a luminosity consistency error. And carrying out self-supervised training on the to-be-trained deep learning model. According to the method, the limitation of high-frequency information loss in a sampling process in a traditional method can be overcome, structural details and boundary information in depth estimation are effectively enhanced and supplemented, and a depth map with higher quality is obtained.

The invention provides an IMU (Inertial Measurement Unit) speed estimation method and system based on multi-scale time sequence feature coding and Mama. The method belongs to the field of IMU speed estimation, and comprises the following steps: S1, acquiring original time sequence data which is generated by an IMU and contains three-axis acceleration and three-axis angular velocity, and intercepting a section of historical time sequence data with a fixed length by taking a current moment as an end point to construct a historical window; s2, inputting the original time sequence data in the historical window into a multi-scale one-dimensional convolution encoder, processing the input data by the encoder by using at least two convolution kernels with different sizes in parallel so as to extract time sequence characteristics under different receptive fields, and fusing output characteristics of each branch to obtain a potential characteristic sequence; and S3, inputting the potential feature sequence into a Mama model to carry out time sequence dependency relationship modeling, and extracting an output state of the Mama model at the end of the sequence as a solved three-dimensional speed output. According to the method, the IMU speed estimation precision can be improved.

The invention provides an estimation method for newly added cross-border mining in a strip mine area based on an unmanned aerial vehicle technology, and relates to the technical field of mining management and resource monitoring. The method comprises the steps of obtaining early-stage and later-stage DEM data and image data, and aligning the DEM data by using bilinear interpolation to obtain matched data; and carrying out pixel-by-pixel difference calculation to generate a grid DEM difference image. And AI interpretation is carried out in combination with the image data, mining area mining range data is obtained, spatial analysis is carried out, and a cross-border mining range is determined. The new border-crossing mining range and the mining quantity estimated value are obtained through mask analysis, the problems that manual inspection and a traditional remote sensing method are low in efficiency and insufficient in accuracy are solved, and real-time and automatic monitoring and estimation of border-crossing mining are achieved.

The invention discloses a depth estimation method and device for any video and a storage medium, and belongs to the field of visual depth estimation. The method comprises the following steps: carrying out annotation processing on a scene video sample to obtain a deep annotation video data set; screening based on the depth labeling video data set and the TartanAir data set to obtain a spatio-temporal joint training sample; the method comprises the following steps: performing time sequence embedding on a multi-head attention layer in an encoder of a DepthAnything model to obtain a space-time combined multi-head attention layer, and constructing an initial TC-DepthAnything model; training the initial TC-DepthAnything model by adopting a space-time joint training sample, and performing constraint by adopting an overall training loss function formed by space consistency loss and time domain regularization loss in the training process to obtain a target TC-DepthAnything model; and inputting any video into the target TC-DepthAnything model to obtain a predicted depth video. The problem of time sequence jitter of DepthAnything in video depth estimation is solved, and flicker artifacts and motion blur in a dynamic scene are inhibited. And video depth estimation with a large application range and an accurate estimation result is realized.

The invention discloses a personalized physical examination period dynamic estimation system and method based on big data, and relates to the technical field of medical health data processing and intelligent evaluation. The estimation method comprises the step of constructing an integrated implementation framework including a multi-source data integration link, a multi-party cooperative processing link, a three-step modeling evaluation link and a quantitative feedback adjustment link. According to the personalized physical examination cycle dynamic estimation system and method based on big data, a medical institution, wearable equipment, family genetic disease history and government affair medical treatment are integrated through a cross-domain data fusion platform, and a multi-dimensional evaluation basis covering physiology, behavior, heredity and resource collaboration is constructed; and then through a three-step modeling process of a hybrid model training engine, the problem of insufficient data is solved through similar group data adaptation, nonlinear correlation features are mined through multiple decision trees, a risk level is finally output, and a period is dynamically adjusted in combination with medical resource assessment.

The invention provides a disturbance estimation method and device based on an observer, equipment and a medium, is used for the technical field of automatic control, and can solve the technical problem of low disturbance estimation accuracy caused by switching impact of the observer in the prior art. The method comprises the following steps: determining observation index values respectively corresponding to a high-order observer and a low-order observer according to corresponding state vectors, and judging whether a target observation index value exists in the observation index values or not; when the judgment result is that the target observation index value does not exist, calculating a high-frequency energy ratio according to the energy values corresponding to the target frequency band and the high-frequency frequency band respectively; substituting the high-frequency energy ratio into the target weight model to obtain a corresponding dynamic low-order weight and a dynamic high-order weight; when the judgment result is that the target observation index value exists, determining a preset low-order weight and a preset high-order weight; performing weighted summation on the disturbance estimated value according to the low-order weight and the high-order weight to obtain a target disturbance estimated value; therefore, the accuracy of disturbance estimation is improved.

The invention discloses a stereo parallax estimation method based on multi-modal feature fusion and generation guidance. The method comprises the following steps: firstly, acquiring a data set composed of binocular image pairs and parallax images; secondly, inputting the binocular image pair into a geometric feature extractor and a semantic feature extractor, and respectively extracting geometric features and semantic features of the binocular image pair; and fusing the geometric features and the semantic features to obtain multi-modal features. And then inputting the multi-modal features into a stereo parallax estimation encoder, and calculating to obtain a predicted parallax map. And finally, based on the predicted disparity map, in combination with the obtained disparity map, training optimization is carried out by using an occlusion perception disparity guide synthesis network. The method shows stronger generalization ability and estimation precision in processing challenging scenes such as weak texture, shielding and illumination variation, and has a wide application prospect.

The invention discloses a partial discharge capacity-severity joint estimation method, a product and equipment, and belongs to the technical field of power equipment monitoring. The method comprises the following steps: S1, constructing a physical fusion framework of multi-modal data: constructing a partial discharge signal propagation graph structure, and converting multi-source heterogeneous data into a unified physical graph; s2, establishing a physical prior model of partial discharge propagation based on graph Laplacian frequency domain features; s3, constructing a correction network to compensate the deviation of the physical model; s4, carrying out capacity-severity joint estimation and uncertainty quantification: carrying out joint modeling on the continuous discharge capacity and the discrete severity level, approximating a real posteriori through a variational posteriori, and outputting point estimation and confidence information; and S5, risk assessment and adaptive optimization: performing risk level assessment based on a result of S4, and performing adaptive adjustment and optimization when a false alarm occurs. The method has continuous signal quantization and discrete grade discrimination capabilities, and is especially suitable for non-contact discharge detection scenes under complex medium conditions.