417 results about "Kernel density estimation" patented technology

The invention discloses a box-type substation state monitoring and early warning method based on artificial intelligence, relates to the technical field of intelligent power grids, and aims to solve the problems of missing report, false report and response lag caused by the fact that an existing static threshold ignores multi-physical coupling and a depth model highly depends on scarce fault samples. According to the scheme, sliding window kernel density estimation is carried out on a multi-channel time sequence signal, a dynamic coupling matrix is constructed through recursion Copula decomposition, a three-level threshold surface is generated through time-varying quantile regression, abnormal samples and graph attention network extraction state representation are generated in combination with a conditional variation auto-encoder, lightweight recursion pruning is carried out, and the dynamic coupling matrix is obtained. An abnormal score is generated through a multilayer Bayesian network and particle filtering, a multi-step risk trend is discriminated through a Gaussian kernel derivative slope, and finally unscented Kalman filtering is used for smoothing and online threshold correction; according to the method, the detection sensitivity and the early warning recall rate of the box-type substation to the transient coupling fault are remarkably improved, the response speed is improved, and the false alarm frequency is effectively reduced.

The invention discloses a three-dimensional scene reconstruction method and device based on large model geometric prior, and a medium, and aims to solve the problems that a conventional 3DGS is liable to have artifacts and detail loss in geometric discontinuity, data redundancy and illumination variation scenes, and predicts a dense depth map and a normal map from a monocular image by using a pre-trained large model. The position and form of the Gaussian kernel are constrained as additional geometric priori; a primitive adjustment strategy based on kernel density estimation is introduced in the training stage, small Gaussian primitives with similar structures and adjacent spaces are combined into a large Gaussian primitive, the rendering quality is kept, redundancy is reduced, and the volume of the model is reduced; an exposure coefficient is adaptively estimated for each input image, an exposure compensation image loss function is constructed, and floating artifacts caused by illumination differences at shooting moments are eliminated. Experiments show that compared with the prior art, the method improves the three-dimensional reconstruction precision and real-time rendering quality of complex illumination and less-texture areas in a public data set and an unmanned aerial vehicle aerial photography scene.

The invention discloses a wind power multi-scale decomposition prediction method. At present, single-point prediction is not comprehensive and accurate enough, and cannot adapt to quantitative accurate requirements of a wind power plant and a power grid dispatching mechanism in risk management. The method comprises the following steps of: forming an original wind power sequence from actually acquired wind power data, sequentially performing feature selection and data decomposition processing to form multi-scale modal data, and constructing a depth prediction model according to the multi-scale modal data; a probability prediction interval determination process is completed in the residual error distribution mode depth prediction model through adaptive bandwidth kernel density estimation; after actually obtained wind power data form an original wind power sequence, an initial model is established, feature selection processing is performed on the initial model, that is, weighted marginal contribution is calculated for each feature of the initial model according to all involved feature subsets by using an SHAP algorithm based on a Shapley value in a game theory, and the weighted marginal contribution of each feature of the initial model is calculated; and completing a feature data acquisition process of accurately quantifying interdependence and interaction effect between features.

The invention relates to the technical field of optical elements, and discloses a fog display point diffusion optical system and a preparation method thereof, and the preparation method comprises the steps: obtaining user visual parameters, and building a visual behavior probability density model through kernel density estimation; designing the front surface of the lens based on the model, and generating an asymmetric defocus system coupled with a gazing habit; the rear surface of the lens is designed, and the distribution density of a fog display point diffusion unit is cooperatively modulated by the front surface defocusing amount, the fixation probability and the pupil diameter; and finally integrating front and rear surface design to form an optical lens body. According to the method, the asymmetric defocus field on the front surface and the fog display point diffusion field on the rear surface are subjected to collaborative design, and the defocus intensity and the contrast modulation intensity are accurately applied to the effective retina area of the user through the visual behavior probability density model; the technical problem that a traditional out-of-focus lens is fixed in signal and cannot adapt to individual staring habits and physiological parameter changes is solved.

The invention discloses a key sensor short-time abnormal distribution drift detection method in a unit start-stop process, and belongs to the technical field of gas turbine power plant financial supervision and artificial intelligence, and the method comprises the steps: synchronously triggering multi-channel signal collection through a main clock, and achieving noise suppression and data pre-screening through the combination of first-order difference and threshold filtering; constructing a nonlinear weighted feature matrix, and fusing a time attenuation coefficient and a shafting acceleration factor to enhance the transient feature expression capability; generating a sensor association graph based on double-threshold determination of weighted Pearson's correlation coefficients and mutual information, and dividing stable subgroups by using an incremental label propagation algorithm; designing a double-layer Cluster-GCN model, aggregating subgroup internal characteristics in the first layer, introducing a fuel valve position-acceleration comparison gating mechanism in the second layer to correct a global edge weight, and generating a node embedding vector sensitive to working condition change; gaussian kernel density estimation and an instantaneous deviation index of embedding similarity are fused, and a historical sliding mean value and subgroup connectivity analysis are combined, so that sensor faults and working condition abrupt changes are distinguished.

The invention relates to the field of power transmission line channel risk hidden danger monitoring, and provides a data fusion power transmission line channel risk hidden danger monitoring method and system, and the method comprises the steps: collecting the multi-modal sensing data of a power transmission line channel, and generating a multi-modal data flow of a unified time-space coordinate; constructing a three-dimensional space point cloud through a phase unwrapping and stereo matching fusion algorithm, and fusing multi-modal data to generate a space probability tensor; extracting risk semantic latent variables, constructing a Bayesian network and identifying potential risks; performing tensor product on the potential risk and the environmental data to generate a dynamic risk enhancement feature matrix, and constructing a nonlinear dynamic threshold curved surface through quantum annealing and Gaussian process regression; a mechanical equation is constructed, Gaussian kernel density estimation and numerical simulation are combined, the evolution trajectory of the risk in the space-time dimension is predicted, and a risk thermodynamic diagram and early warning information are generated; and generating a structured risk early warning report by adopting a natural language processing method. And the accuracy of power transmission line channel risk hidden danger monitoring is improved.

The invention discloses a sudden death risk real-time assessment system and method based on multi-modal physiological signal fusion, relates to the field of human physiological state monitoring and early warning, and solves the problems of low accuracy and poor real-time performance of sudden death risk assessment by single-modal signals. The system comprises a signal acquisition module, a preprocessing module, a high-dimensional feature extraction module, a multi-modal feature fusion module, a sudden death risk quantification module, a model updating engine module and an early warning feedback module, and each module integrates a self-adaptive filtering unit, a time sequence convolutional network unit, an improved multi-head self-attention mechanism unit and the like. According to the scheme, multi-mode signals such as electrocardio are synchronously collected through multiple channels, a dynamic risk index is calculated through self-adaptive noise reduction, parallel extraction of time-frequency domain nonlinear features and mutual information weighted fusion in combination with kernel density estimation, and online incremental updating and multi-stage early warning of a model are achieved; the sudden death risk can be accurately evaluated in real time, the anomaly detection sensitivity and the early warning timeliness are improved, and the method is suitable for daily health monitoring and high-risk group risk management and control.

The invention discloses an urban rail transit pull-in passenger flow prediction method fusing multi-source spatio-temporal data. The method comprises the following steps: collecting multi-source spatio-temporal data of a target station and a corresponding associated station; kernel density estimation is carried out on the surrounding POI density corresponding to each site, the spatial thermal characteristics of each site are generated, one-hot coding is carried out on the real-time weather index corresponding to each site, the weather influence characteristics of each site are generated, segmented coding is carried out on the date type identifier corresponding to each site, and the periodic effect characteristics of each site are generated. Performing differential stabilization on the historical pull-in passenger flow sequence of each station to generate a time sequence fluctuation characteristic of each station; tensor splicing is carried out on the space thermal characteristics, the weather influence characteristics, the periodic effect characteristics and the time sequence fluctuation characteristics of all the stations, a space-time fusion characteristic tensor is constructed and input into a space-time self-adaptive prediction model, and an inbound passenger flow prediction value of the target station is generated. The accuracy and timeliness of urban rail transit pull-in passenger flow prediction can be improved.

The invention discloses a power system random scheduling method based on wind-solar joint probability distribution and double-layer dynamic optimization. The method comprises the following steps: firstly, constructing a wind and light output joint probability distribution model by adopting nonparametric kernel density estimation and a Frank Copula function so as to accurately describe space-time correlation; secondly, generating a wind-solar combined output scene through Monte Carlo simulation, and performing reduction by adopting a k-means + + clustering algorithm to obtain a typical scene set; secondly, establishing a double-layer optimization model, wherein the upper layer takes the minimum net load variance as a target to coordinate wind and light storage output stabilizing fluctuation; and the lower layer optimizes the output of the traditional unit by taking the minimum operation cost of the system as a target. Finally, a dynamic climbing constraint mechanism is introduced, the net load fluctuation standard deviation output by the upper layer is converted into a dynamic constraint threshold value of the climbing rate of the lower layer unit, collaborative optimization of economical efficiency and stability is achieved, and the dispatching robustness of the high-proportion renewable energy power system is remarkably improved.

The invention discloses a power distribution network photovoltaic openable capacity dynamic evaluation method fusing voltage stability margin and neural network optimization, and aims to solve the problems that a traditional method does not fully consider dynamic stability and is low in calculation efficiency. Firstly, a prediction method combining kernel density estimation and quantile regression is adopted to accurately quantify the uncertainty of distributed photovoltaic output. The core innovation of the method is that on the basis of traditional security constraints, a static voltage stability margin (VDSM) is introduced as a key constraint condition, and a double-layer interval analysis model capable of guaranteeing the dynamic stability of a power grid is constructed. Secondly, in order to efficiently solve, the invention provides a framework of'neural network pre-screening + parallel optimization ': after a model is decomposed into optimistic sub-problems and pessimistic sub-problems through an interval decoupling technology, massive candidate solutions are quickly screened by utilizing a neural network model, so that a feasible solution space is greatly reduced, and the solution efficiency is improved; and carrying out parallel optimization solution on the sub-models in combination with an improved particle swarm optimization algorithm.

The invention relates to the technical field of power system planning, and discloses a new energy grid-connectable capacity margin evaluation method based on spatio-temporal distribution characteristics, which comprises the following steps: constructing a joint probability distribution model based on nonparametric kernel density estimation and a Copula function by using historical and meteorological data of a wind and light station, and generating a random output scene reflecting the spatio-temporal complementarity of wind and light resources; constructing a multi-dimensional stability constraint set containing a generalized short circuit ratio and broadband oscillation as a safety boundary of a random output scene; establishing a grid source interaction control strategy considering energy storage adjustment and direct current power modulation by taking grid-connected capacity maximization as a target, solving an optimal installed capacity scheme, and activating the grid source interaction control strategy to perform dynamic correction when a scene is out of limit; the maximum grid-connected capacity distribution meeting the constraint is counted, and the shadow price of the constraint is calculated to identify the bottleneck. According to the method, the wind and light resource complementary benefit and the system regulation potential can be quantified, and the new energy grid-connected capacity evaluation margin is improved on the premise of guaranteeing the safety.

The invention discloses a UWB (Ultra Wideband) ranging error suppression method fusing a black-wing plinary algorithm and an information entropy adaptive adjustment mechanism. The method comprises the following steps: firstly, identifying a non-line-of-sight signal by combining a ranging residual error, a first-order difference and a sliding window strategy; abnormal values are removed based on residual errors, missing data are compensated through segmented three-time Hermite interpolation, and data continuity and smoothness are guaranteed; calculating Shannon information entropy by using sliding window kernel density estimation, dynamically quantifying residual distribution uncertainty, adaptively adjusting Kalman gain, and realizing weight real-time adjustment; meanwhile, the initial parameter configuration of the filter is optimized by adopting a black-wing algorithm, and the convergence and the overall stability of adaptive filtering are improved. Aiming at the problems of measurement abnormity and precision reduction caused by the fact that UWB distance measurement is easy to be shielded and interfered in the NLOS environment, the method can effectively inhibit distance measurement errors, overcomes the limitation that a traditional method depends on manual parameter adjustment, and provides powerful technical support for high-precision UWB distance measurement in the NLOS environment.

The invention discloses a 3D GS cultural relic digital reconstruction method and system based on a block chain, and the method comprises the steps: collecting the RGB image data and depth perception data of a cultural relic, eliminating the influence of different shooting conditions through an illumination separation processing technology, and building a standardized image data set; recognizing a surface area suitable for reconstruction based on image analysis, determining feature point distribution by using kernel density estimation, and generating initial three-dimensional representation through Gaussian ellipsoid fitting; performing gradient calculation and feature extraction on the depth data, and combining with Gaussian representation to form a geometric constraint mechanism; self-adaptive encryption based on visual importance is executed for a sparse region, and a layered rendering effect is achieved through opacity parameter adjustment; the rendering characteristics and the conversion relation of different view angles are analyzed, key observation points are determined through stability analysis, and a smooth multi-view-angle display sequence is constructed; and integrating multi-view rendering information to generate volumetric representation, and completing right confirmation of the high-quality three-dimensional digital model through digital signature.

The invention relates to the technical field of intelligent inspection, in particular to a solid waste storage yard remote cruise and hidden danger recognition system based on the AI technology, which comprises a multi-source monitoring data acquisition module, a storage yard feature enhancement processing module, a multi-mode hidden danger collaborative recognition module and a dynamic cruise path generation and execution module. The unmanned aerial vehicle carries a multispectral camera and a laser radar to cruise according to an initial path, collects visible light images, infrared images and terrain point cloud data of a storage yard, and receives leachate monitoring data. Image enhancement is realized through optical dust scattering compensation and a pile body surface curvature mapping graph; the deformable convolution unit is combined with percolate temporal and spatial change rate anomaly judgment to trigger high-precision identification, and outputs a hidden danger type and a space coordinate set; according to the method, a closed loop of recognition, path optimization and re-acquisition is realized, and the hidden danger detection precision and the monitoring efficiency are improved.

The invention discloses a tracking optimization method for detecting a low-speed small unmanned aerial vehicle target by a single-photon laser radar, belongs to the technical field of optical detection and target tracking, and aims to solve the problem of unstable target imaging and tracking caused by speckle noise interference in the prior art. The invention provides a speckle noise suppression method based on a vibration emission optical fiber and a space-time dynamic kernel density estimation algorithm, and the method is combined with an improved mean shift-Kalman filtering algorithm to achieve the tracking optimization of a low-speed small unmanned aerial vehicle target, and obtains an echo signal through a single-photon laser radar. Constructing a three-dimensional data matrix and reconstructing distance and intensity images through space-time filtering and kernel density estimation; and then, in combination with gray level histogram probability estimation, similarity measurement and Mean Shift iteration, target area tracking is realized, and finally, a target position is predicted and output by using Kalman filtering. The method is suitable for the fields of long-distance unmanned aerial vehicle detection and monitoring, low-altitude security defense early warning, civil airspace management and control, military anti-unmanned aerial vehicle systems and the like.

The invention provides a large model training data desensitization protection method and system based on differential privacy, and relates to the technical field of artificial intelligence security, and the method comprises the steps: calculating a differential privacy parameter, mapping training data to a feature subspace, and constructing a dimension reduction manifold projection operator; solving an optimal disturbance vector by using a Lagrange multiplier method to form a de-sensitization feature set; probability distribution is constructed based on kernel density estimation, and Gaussian noise is injected; and dynamically adjusting a parameter group updating strategy in model training. According to the method, data utility and privacy protection can be effectively balanced, the large model training safety is enhanced, and the model convergence efficiency is improved.

The invention discloses an HE event early warning method, device and equipment, a medium and a product, and relates to the field of mine earthquake early warning. The method comprises the following steps: analyzing space-time correlation of mine earthquake data energy by adopting a semi-variation function according to acquired mine earthquake monitoring data, and determining a historical data window; constructing a dynamic sliding mechanism; based on a dynamic sliding mechanism, adopting a principal component analysis method and a kernel density estimation method to extract a kernel density peak value of the micro-seismic event, and adopting a fractal dimension analysis method to determine a fractal dimension index so as to analyze and quantify the geometric complexity change condition and obtain a quantification result; based on a quantification result, optimizing early warning parameters by adopting grid search, and determining a sliding window joint early warning model; and monitoring a kernel density peak value and a fractal dimension index in real time by adopting a sliding window combined early warning model based on the optimized early warning parameters so as to realize early warning of the HE event. The invention aims to improve the accuracy and timeliness of HE event prediction.

The invention discloses a pollutant tracing method and system based on intelligent fingerprint database matching. The method comprises the following steps: acquiring historical data of each process production area of a target industrial gathering area, and constructing a fingerprint database containing pollutant types, concentrations and change information; calculating the SHAP value of each specific pollutant by using a machine learning model to serve as an initial weight, constructing a correction coefficient in combination with environment durability and detection frequency, and fusing to obtain a specific weight vector; based on the weight, constructing a multi-modal mixed kernel density estimation model of each emission area under multiple processes by adopting a kernel density estimation method; the matching degree of each pollution source and the sample is determined by calculating the similarity score of the to-be-traced sample and the pollutant concentration distribution in the atlas database, so that the accurate and rapid traceability of the pollution emission area is realized. According to the invention, the accuracy and adaptability of pollutant traceability in a complex industrial environment are effectively improved.

The invention discloses an electric power and electric quantity balance analysis scene generation method and system considering source load uncertainty, and the method comprises the steps: carrying out the fitting of the distribution of new energy output and load in each time period through employing kernel density estimation according to the historical data of new energy and load, so as to obtain an uncertainty model of a source load prediction error; historical net load data is calculated based on source load historical data, a K-means clustering algorithm is adopted to cluster net load curves, typical net load curve types are obtained, and the occurrence probability of each type is calculated; on the basis of different types of net load curves, according to the new energy and load proportion superposition source load prediction error uncertainty, obtaining a typical scene of power system operation under the high-proportion new energy; and constructing an optimization model with the aim of minimizing the power and electric quantity balance gap of the local power grid / maximizing the new energy consumption, and analyzing the power and electric quantity balance problem of the local power grid researched under the access of the high-proportion new energy by taking the typical scene as input. The method can efficiently and reasonably construct the operation scene of the power system.

The invention provides a non-stationary industrial process monitoring method and system. The method comprises an offline training stage: calculating a time Laplacian matrix and a space Laplacian matrix based on a historical data matrix; constructing an objective function of the stationary subspace analysis method, and adding a time constraint term of a time Laplacian matrix and a space constraint term of a space Laplacian matrix into the objective function; solving the objective function to obtain a stable projection matrix; calculating a stationary component and a monitoring index of each sample in the data matrix X in sequence; determining a control limit by using a kernel density estimation method; an online monitoring stage: based on the real-time operation data x, calculating a stationary component of the real-time operation data x and a corresponding real-time monitoring index according to the stationary projection matrix, and if the real-time monitoring index is greater than a control limit, judging that the operation of the non-stationary process has a fault; the monitoring accuracy can be improved.

The invention relates to a dynamic density guided generative adversarial network unbalanced data processing method (DAG-WGAN). The DAG-WGAN realizes unbalanced data processing through data preprocessing, dynamic density estimation and weight distribution, potential structure learning based on a variational auto-encoder (VAE), and density guide generation and dynamic feedback optimization based on WGAN-GP. The DAG-WGAN adaptively evaluates the sample density by using kernel density estimation (KDE) and a Gaussian kernel function, and allocates a weight for a generation process, thereby emphatically enhancing the low density and discriminating the sample generation of a difficult region. The VAE learns a potential manifold structure of a minority class of samples, realizes density-guided generation of a potential space under a WGAN-GP framework, and ensures diversity and manifold consistency of generated samples. In addition, a dynamic feedback mechanism is introduced, the weight and the gradient penalty coefficient are adaptively adjusted and generated, and the training stability and the sample generation robustness are improved.

The invention discloses a photovoltaic power prediction method based on photovoltaic multi-scene generation and interpretability, and the method comprises the steps: generating a combined output scene of a plurality of photovoltaic units through kernel density estimation and an improved Copula function, providing more precise scene simulation, employing an improved ISODATA clustering method to improve the precision of scene clustering, and carrying out the prediction of the combined output scene. Therefore, the uncertainty characteristics of photovoltaic power generation can be described more comprehensively; then, a Stacking regression method is adopted, and the reliability and stability of prediction are improved by integrating multiple base learners; and finally, in combination with an SHAP interpretability analysis method, the decision process of the model and the importance of each feature in prediction are deeply understood, and the transparency of the model and the interpretability of the decision process are enhanced. According to the method, the accuracy of photovoltaic prediction can be effectively improved, enough transparency and trust are provided for users, and the method has certain engineering use value.

The invention relates to the technical field of electric power transaction data processing, and discloses an electric power spot transaction data time sequence mining method capable of representing market fluctuation characteristics, which comprises the following steps: executing variational mode decomposition on electricity price data, and locking a probability density trough by using kernel density estimation to determine a dynamic shunting threshold value; according to the method, self-adaptive boundary locking in a non-stationary environment is realized by utilizing a data topology form, invalid feature interference is blocked through mutual information causal gating, and the adaptive boundary locking in a non-stationary environment is realized by utilizing a linear extrapolation model and a multi-head attention mechanism in which a causal mask is introduced. The contradiction between mean regression and peak capture of a traditional model is solved, and the effectiveness of a prediction result is improved.

The invention discloses a progressive forecasting method for regional persistent rainstorm based on key impact factors, which comprises the following steps of: firstly, comparing and analyzing difference characteristics of a regional persistent rainstorm process and a general rainstorm process in a traditional physical quantity factor, a physical quantity anomaly factor and a comprehensive dynamic factor, and screening a key impact factor set; then, extracting a circulation field anomaly factor before rainstorm occurrence to construct a quantitative index system, and establishing a three-level linkage criterion in combination with a mid-high weft circulation form, a mid-low weft power lifting threshold value and water vapor conveying strength; and finally, determining an optimal factor combination and weight distribution scheme by taking the TS score as a judgment criterion, constructing a probability forecasting model by fusing a kernel density estimation technology, and outputting a rainstorm falling area spatial distribution probability. According to the method, a traditional single forecasting mode is broken through, a progressive forecasting technology system from rainstorm process forecasting to falling area probability forecasting is innovatively established, and technical support is provided for early warning decision making and disaster prevention and reduction deployment of extreme weather events.

The invention belongs to the technical field of new energy station power prediction, and particularly relates to a new energy station power interval prediction method. In order to overcome the defect that the prediction accuracy and precision of an existing new energy station power interval prediction method have the improvement space, the invention adopts the following technical scheme: the new energy station power interval prediction method comprises the following steps: constructing a VMD-SA-TCN-BiLSTM-KDE prediction model; historical output power data of a wind power plant station are obtained, the historical output power data are preprocessed and then input into SA-TCN-BiLSTM for training, and the preprocessing comprises the step of decomposing the data through VMD; the part of historical output power data obtained after preprocessing is input into the trained SA-TCN-BiLSTM, and a prediction result is obtained; and performing KDE kernel density estimation on the prediction error, and obtaining a historical output power prediction interval of the wind power plant station according to a prediction result and the prediction error. The method has the beneficial effect of higher prediction precision.

The invention discloses a new energy uncertainty interval fine modeling method based on ELM-Copula. Comprising the following steps: 1) data reading and preprocessing: obtaining a power prediction value and an actual output value by reading historical operation data of a new energy electric field, performing kernel density estimation on the per-unit power prediction value and the actual output value, and calculating an error absolute value according to the per-unit power prediction value and the actual output value; 2) constructing a dynamic Copula function model; 3) measuring goodness of fit of the model; 4) calculating a confidence interval of a prediction error, analyzing uncertainty of power generation power prediction, and giving a confidence interval of a prediction value; according to the method, a multi-type dynamic Copula model is introduced to construct a dynamic dependency structure between a prediction error and power, the ELM is applied to a post-processing stage of a dynamic Copula prediction interval, a correction coefficient is generated by learning historical deviation characteristics, and an original interval is shrunk, so that the prediction precision and practicability are improved on the premise that the coverage rate is not reduced.

The invention relates to a spacecraft maneuver detection method based on TLE data and kernel density estimation. The invention relates to the technical field of spacecraft maneuver detection. The method comprises the following steps: acquiring and inputting satellite historical TLE data; gaussian filtering denoising is carried out on the input data; based on the de-noised data, propagation forecasting is carried out through an SGP4 model; constructing a joint residual sequence, and carrying out error distribution modeling and sample scoring; determining a maneuvering threshold value, and judging to obtain abnormal data; and aggregating abnormal data and carrying out maneuvering identification. The method can still realize accurate and stable track maneuver automatic detection under the constraint conditions of unstable TLE data, complex error distribution, no external auxiliary data and the like. The method has good adaptability, robustness and interpretability, and is a track anomaly detection technical scheme which is complete in structure, rigorous in logic and capable of achieving engineering landing.

The invention relates to the technical field of intelligent operation and maintenance of an electric power system, in particular to an inspection sample image data enhancement method for an electric power artificial intelligence platform, which is used for solving the problems that in the prior art, history and equipment knowledge cannot be fused to construct a forbidden area, a co-occurrence rule and component association, cross-component defect positions cannot be effectively adjusted, and the detection accuracy is poor. Defect distribution is difficult to accurately control, and physical rationality and engineering credibility are reduced. According to the method, a forbidden area, a co-occurrence rule and component association are constructed by fusing history and equipment knowledge, masks are generated through kernel density estimation to suppress invalid defects, co-occurrence frequencies are counted based on feature vectors, co-occurrence relationships are determined by combining distances and similarities, and the masks, matrixes and graphs are embedded and coded into conditional vectors, so that the non-ineffective defects are suppressed. And zero setting is performed on a forbidden area in the generative network, illegal co-occurrence is filtered, and cross-component defect positions are adjusted, so that defect distribution is accurately controlled, and physical rationality and engineering credibility are enhanced.

The invention discloses a GIS partial discharge fault positioning method based on multi-feature fusion and kernel density estimation optimization, and the method comprises the following steps: collecting a partial discharge original pulse signal, carrying out the signal processing, and outputting a first processing signal; performing adaptive robust baseline determination on the first processing signal, detecting a wave head position based on multi-feature fusion and a dynamic threshold voting mechanism, and outputting a wave head arrival time; carrying out time synchronization matching on the wave head arrival time, and calculating and outputting a preliminary positioning result based on a time difference of arrival method; and accumulating a plurality of preliminary positioning results, performing optimization by adopting a kernel density estimation method, and outputting a final partial discharge fault position. According to the method, impulse noise and white noise are effectively suppressed, stable signals are provided for subsequent processing, the detection precision of the wave head arrival time is remarkably improved, the problem that false detection and missing detection are caused by the fact that a traditional fixed threshold value or a single feature is prone to noise interference is solved, the random error of single-time positioning is greatly compressed to be within 0.07 m from 0.33 m on average, and the overall positioning precision is improved by 80%.

The invention relates to a modeling method for accurate traffic flow prediction, which comprises the following steps of S1, integrally modeling all intersections and roads into a road network undirected graph, and defining an adjacent matrix and a distance matrix, S2, defining propagation time delay of traffic flow at the intersections, and S3, calculating the traffic flow of the intersections according to the propagation time delay. S3, combining propagation time delay and traffic flow abrupt change influence of adjacent intersections to confirm a comprehensive effect of each intersection at the moment t, and obtaining traffic flow representation of the intersection at the moment t, S3, aggregating features of each intersection through an adjacent matrix and the propagation time delay by a space graph convolutional layer, and capturing a dynamic rule through a time graph convolutional layer, the method comprises the steps of (S1) obtaining a graph convolution layer, combining the graph convolution layer and a time convolution layer to form a space-time graph convolution network, and outputting final traffic flow prediction, (S2) giving the final traffic flow prediction and a kernel density estimation matrix of GKDE and setting three learnable matrixes, and (S5) outputting through a feedforward neural network. The method has the advantage of accurate traffic flow prediction.