3585 results about "Data point" patented technology

Multi-modal models learn a joint latent space for relating data points across different modalities. To more effectively learn multi-modal models with reduced training requirements and greater benefit from limited multi-modal training data, a multi-modal model may be trained with fixed or pre-trained unimodal encoders that generate data representations in respective latent spaces. The multi-modal model is trained to learn a shared latent space while fixing the unimodal encoders, enabling training without storing the unimodal encoders in memory. Limited multi-modal data may also be augmented by generating synthetic data between commonly-labeled pairs in the respective modality's latent spaces. The effect of data diversity can also be determined by generating a diverse data set with respect to the data points in latent space, enabling measurement of performance of the multi-modal model on limited training data.

Multi-modal models learn a joint latent space for relating data points across different modalities. To more effectively learn multi-modal models with reduced training requirements and greater benefit from limited multi-modal training data, a multi-modal model may be trained with fixed or pre-trained unimodal encoders that generate data representations in respective latent spaces. The multi-modal model is trained to learn a shared latent space while fixing the unimodal encoders, enabling training without storing the unimodal encoders in memory. Limited multi-modal data may also be augmented by generating synthetic data between commonly-labeled pairs in the respective modality's latent spaces. The effect of data diversity can also be determined by generating a diverse data set with respect to the data points in latent space, enabling measurement of performance of the multi-modal model on limited training data.

Multi-modal models learn a joint latent space for relating data points across different modalities. To more effectively learn multi-modal models with reduced training requirements and greater benefit from limited multi-modal training data, a multi-modal model may be trained with fixed or pre-trained unimodal encoders that generate data representations in respective latent spaces. The multi-modal model is trained to learn a shared latent space while fixing the unimodal encoders, enabling training without storing the unimodal encoders in memory. Limited multi-modal data may also be augmented by generating synthetic data between commonly-labeled pairs in the respective modality's latent spaces. The effect of data diversity can also be determined by generating a diverse data set with respect to the data points in latent space, enabling measurement of performance of the multi-modal model on limited training data.

A system and method are provided for examining data from a source. The method is executed by a device having a processor and includes receiving a set of historical data and a set of current data to be examined, from the source. The method also includes generating multiple statistical models based on the historical data and a forecast for each model. The method also includes selecting one of the multiple statistical models based on at least one criterion, and generating a new forecast using the selected model. The method also includes comparing the set of current data against the new forecast to identify any data points in the set of current data with unexpected values. The method also includes outputting a result of the comparison, the result comprising any data points with unexpected values.

The invention provides a valve actuator dispatching control method and system based on AI intelligence, and relates to the technical field of industrial automation control, and the method comprises the steps: collecting the running state parameters of a steam turbine in real time, determining the relation between the steam inlet amount and the valve opening degree under the current working condition to obtain a preliminary valve opening degree estimated value, and carrying out dynamic response compensation; the method comprises the following steps: carrying out comprehensive coupling analysis on a theoretical valve opening value and mechanical parameters such as response time, inertial parameters and an opening limit threshold value of a valve actuator, calculating a dynamic adjustment weight and carrying out boundary constraint processing to obtain a processed parameter, and calling a preset valve opening-silicon controlled trigger angle nonlinear mapping model in a single chip microcomputer based on the parameter; and calculating a final silicon controlled trigger angle by inquiring adjacent data points and a linear interpolation method. By fusing intelligent prediction, multi-dimensional adjustment and a closed-loop feedback mechanism, the valve control precision and response speed are improved, and efficient and intelligent operation of industrial automatic control is achieved.

An approach to time-series data point anomaly detection may be presented. Data point anomalies in time-series data can cause a cascade of incorrect predictions in a time-series data prediction model. Presented herein may be an approach to decompose a time-series training data set into elementary components, such as seasonal, trend and residual. The approach may determine one or more confidence intervals for elementary components of data points including level shift, variance, and outlier. From these confidence intervals, new data points can be analyzed and identified as anomaly data points. The approach may also prevent anomaly data points from being incorporated into a time series data prediction model, reducing prediction error in the prediction model.

A system for the temporal ordering of multi-channel data streams, comprising the following: a multi-channel data reception module configured to receive real-time data streams from multiple sources, extract temporal metadata, and detect discrepancies in timestamps; a time synchronization module configured to align data packets using timestamp normalization, adaptive buffering, and AI-based predictive modeling; an AI-powered predictive modeling module trained to estimate missing or delayed timestamps and to reorder out-of-order data packets based on learned patterns; an adaptive buffer module configured to dynamically adjust buffer sizes based on real-time data arrival rates and network conditions; an ordering and output processing module configured to reconstruct the correct order of the data packets and transmit them to a processing or analysis system; and a module for error detection and correction that is configured to detect inconsistencies in timestamps, transmission delays and missing data points, thus ensuring increased reliability of the ordered data sequence.

The invention relates to the technical field of big data, and discloses a space-time big data abnormal event detection method and device and a storage medium. The method comprises the following steps: performing timestamp unification and space coordinate standardization processing on urban space-time big data to obtain a standardized space-time data set; feature vector cosine similarity between each data point in the standardized spatio-temporal data set and other data points is calculated to obtain a target similarity value, and an initial neighborhood data set is selected; performing space-time density distribution calculation on the initial neighborhood data set to obtain a core neighborhood data set; performing time sequence feature and spatial distribution feature extraction on the core neighborhood data set to obtain a fusion feature weight matrix; and inputting the fusion feature weight matrix into a support vector machine classifier to perform abnormal event recognition to obtain an abnormal event detection result, thereby improving the classification capability of high-dimensional nonlinear features, and simultaneously realizing all-directional feature description and detection of abnormal events.

The invention relates to a steel structure component virtual pre-assembly method based on three-dimensional laser scanning. Designing a self-adaptive hierarchical scanning strategy, and dynamically adjusting parameters of scanning equipment to obtain data of each region of the component; abnormal points are detected and eliminated by fusing scanning data of different levels; datum points are extracted, coordinate calibration is carried out, data of all levels are spliced, and a full-component three-dimensional point cloud model is formed; classifying the point clouds by using a multi-dimensional data analysis algorithm; establishing a closed-loop data verification system; designing a multi-feature point matching algorithm; performing iterative docking and position calibration through an algorithm; quantitative analysis is carried out on spatial deviation between the components in the assembling process; after the deviation is detected, a fine adjustment scheme is generated, and the position of the component is adjusted; a virtual pre-assembly path is generated according to the shapes, the connection modes and the installation sequence of the components, and the assembly sequence and the layout of spatial positions are controlled; and the assembly state of each component is monitored in real time, and the path is optimized and adjusted through a feedback mechanism.

The invention discloses an intelligent definition method and system for industrial edge data acquisition, a medium and equipment. The method comprises the following steps: acquiring operation data of field equipment in real time; performing protocol analysis and cleaning on the data to generate standardized data points, dynamically detecting the communication state and adaptively adjusting the acquisition frequency; classifying and aggregating the standardized data points into a running state feature set and calculating feature indexes; edge side anomaly detection is carried out based on the feature set, and a protocol container library is synchronously called to match a device communication protocol to generate matching information; an acquisition strategy optimization instruction is generated in combination with the abnormal result and the matching information, and transmission characteristic indexes and instructions are packaged; and structuring storage feature indexes according to equipment types and time dimensions to form a historical operation database, and managing a storage period by adopting a sliding window mechanism. According to the invention, protocol adaptive analysis, dynamic acquisition adjustment and edge intelligent analysis are realized through software definition, the hardware dependence and field debugging risk are reduced, and the data acquisition efficiency and the intelligent level are improved.

The invention relates to the technical field of data analysis, in particular to an intelligent hydrological data analysis method and system, and the method comprises the following steps: collecting water level, flow velocity and precipitation data, calculating the change rate of adjacent moments, marking hydrological parameter abrupt change points, carrying out the statistics of time intervals, predicting the duration of extreme events, and obtaining the distribution characteristics of the hydrological abrupt change points. According to the method, by analyzing the time interval and the duration time of the data points, the distribution characteristics of hydrological abrupt change can be mastered in the early stage, response can be made in advance, the propagation sequence and the diffusion rate of the abrupt change points in different river reaches can be analyzed, and accurate division of the influence range and the risk level of hydrological events is enhanced; effective management of hydrological event space risks is achieved, understanding of hydrological key influence factors is deepened by screening key variables and evaluating the evolution relation of the variables along with time, accuracy of short-term and long-term hydrological risk prediction is enhanced, and a scientific basis is provided for dynamic early warning of hydrological abnormal events.

The invention discloses a multi-modal fused intelligent question answering and knowledge retrieval method and system, and the method comprises the steps: building a multi-modal data index model oriented to a heterogeneous knowledge source, carrying out the feature mapping of text, image, table, chart, audio and video contents through a unified semantic embedding space, and generating a cross-modal index set; after a query request is received, performing semantic matching and structure matching on the cross-modal index set by using a multi-channel retriever to obtain candidate evidence fragments; and based on an evidence granularity decomposition strategy, performing minimum evidence unit division on text statements, table units, chart data points and multimedia frame contents in the candidate evidence fragments, and establishing a semantic consistency graph among the units. According to the method, high-credibility traceable generation of question and answer results is realized through multi-modal fusion and space-time consistency constraint, and the retrieval precision and interpretation transparency in a complex knowledge scene are remarkably improved.

The invention relates to the technical field of calibration tests, in particular to an intelligent metering calibration method and system, and the method comprises the following steps: calculating the similarity between data points, constructing a similarity matrix, further calculating the fitting degree, carrying out the spectral decomposition, extracting a feature vector, carrying out the dimension reduction, mapping to a low-dimensional space, carrying out the data point clustering based on a feature vector matrix, and carrying out the calibration of the data points. An error cluster is identified and an error pattern is determined. According to the method, by calculating the similarity between the data points and mapping the similarity to the low-dimensional space, the error modes can be efficiently recognized and classified, the error correction path can be dynamically adjusted, error correction is more flexible and real-time, the error analysis depth is improved through spectral decomposition and feature vector extraction, and the error analysis accuracy is improved. Therefore, the error trends of the equipment under different conditions are accurately identified, the measurement precision and stability are remarkably improved, the error correction path parameters are updated in real time, the adaptability of the system to the operation change of the equipment is enhanced, and the calibration time is effectively shortened.

The invention provides an intelligent network connection automobile abnormity detection system and method based on big data analysis, and the method comprises the steps: obtaining metadata in the communication between a vehicle CAN bus and a cloud, the metadata comprising a protocol type and a load feature, carrying out the analysis of the load feature according to the protocol type, and extracting a key field of a data package; constructing a load feature matrix according to the key field, and normalizing the heterogeneous data to obtain a feature data set in a uniform format; a mahalanobis distance algorithm is adopted to calculate the distance between each data point in the feature data set and a preset base line, and if the distance exceeds a preset threshold value, abnormal traffic is judged, and an abnormal traffic mark list is generated; fingerprint features of a transmission link are obtained from the 5G network slices, the fingerprint features of different slices are subjected to standardization processing, and a cross-slice consistent fingerprint feature data set is generated; therefore, security threats in a complex Internet of Vehicles environment can be effectively handled, and the accuracy and reliability of abnormal traffic detection are improved.

The invention discloses an abnormal behavior detection method and system based on a multi-source log and electronic equipment, and the method comprises the steps: collecting log data, carrying out alignment processing and data preprocessing, and generating normalized log data; carrying out attribute screening on the standardized log data set, and packaging the screened logs into a JSON format; extracting features of the log data, converting the features into data points, clustering the data points, and retaining a log cluster according to a clustering result; mining a hidden association mode between log attribute items in the log cluster by using an improved Apriori algorithm, generating an association rule, and constructing a knowledge graph based on the association rule; and constructing a time sequence of the log data by using a self-adaptive time window, and inputting the time sequence and the tag sequence of the log data into the LSTM neural network by combining context information in the knowledge graph and a tag sequence generated by an association rule to obtain an abnormal behavior detection result. According to the method, the log data format is unified, correlation analysis is enhanced, and the accuracy of an anomaly detection result is obviously improved.

The invention provides a mangrove forest restoration area carbon sink quantity data monitoring system, and relates to the technical field of ecological carbon sink metering, and the system comprises an acquisition module which is used for collecting a carbon exchange original data set of an atmosphere-vegetation-soil interface based on a multi-type sensor network arranged in an intertidal wetland of a mangrove forest restoration area; the data generation module is used for carrying out space-time coordinate matching based on the original data set and generating a multi-dimensional carbon flux fusion data set of geographic position and time joint indexing; and the identification module is used for calculating a local space-time dispersion and a global correlation threshold value based on the fused data set, and identifying and marking abnormal data points. According to the invention, real-time and multi-dimensional monitoring of the carbon sink amount of the mangrove forest restoration area is realized, the monitoring efficiency and the data reliability are improved, and support is provided for assessment of restoration and management strategies.

The invention, which relates to the technical field of landslide monitoring, discloses an optical fiber sensing landslide dynamic response monitoring system comprising an event monitoring module, a sampling optimization module, a landslide risk analysis module and an early warning level adjustment module. By comparing the offset of adjacent data points in real time, tiny displacement abnormity can be quickly identified, instant marks are provided for early geological activities, and the sampling frequency is automatically adjusted according to real-time monitoring data, so that data acquisition is finer, meanwhile, storage and processing of irrelevant data are reduced, the operation efficiency of the system is effectively improved, and the working efficiency of the system is improved. Real-time data and a historical landslide mode are finely compared, comprehensive analysis is carried out in combination with key indexes such as earth surface displacement and soil humidity, the system can accurately calculate the probability of occurrence of the landslide, the accuracy and timeliness of early warning are greatly improved, and casualties and property losses caused by landslide disasters are effectively reduced.

The invention discloses a welding seam track extraction method and system based on RANSAC parameter fitting, and the method comprises the steps: S1, obtaining three-dimensional point cloud data, and obtaining the normal vector and local curvature of each point cloud data point; selecting an initial seed point from the three-dimensional point cloud data according to the local curvature of each point cloud data point; s2, establishing a feature similarity evaluation system, and adding the initial seed points and the corresponding similar adjacent points into the same plane; according to different planes where the point cloud data points are located, plane area coarse segmentation is carried out; and S3, a plurality of geometric fitting models are constructed for the welding track based on the RANSAC algorithm, the geometric fitting model with the highest matching degree is selected from the geometric fitting models, and the welding track in the welding area is extracted. According to the technical scheme, the welding seam track can be rapidly and effectively extracted, the universality of welding seam track recognition is improved, and the extraction error of the welding seam track is remarkably reduced.

The invention discloses a PINNs high Reynolds number flow field solving method based on regulatory factor optimization, and the scheme comprises the steps: firstly building a physical model, on one hand, building a geometric model according to the actual condition of a reactor core rod bundle channel, and on the other hand, building a fluid mechanics model containing continuity, momentum conservation and energy conservation equations; a data set is obtained, and experimental data is preprocessed and then integrated with simulated data through experimental measurement and numerical simulation; and then training a physical information neural network, designing a neural network model of specific input and output, introducing a regulatory factor to balance the difference between a physical equation and data, constructing a comprehensive loss function containing internal data points, boundary data points and equation loss, and training the network by using a training data set until convergence. And finally, solving a flow field by using the trained neural network, outputting predicted values of a velocity field, a pressure field and a temperature field, and solving the high Reynolds number flow field of the reactor core.

The present disclosure relates to a method and apparatus for detecting anomalous shaking of a vehicle, a device, a vehicle, a medium, and a program product, the method comprising: acquiring a target time sequence corresponding to an original time sequence signal, the original time sequence signal being a time sequence signal corresponding to a vehicle status signal, and the target time sequence being a stable time sequence obtained from preprocessing the original time sequence signal; randomly selecting a plurality of sample points from the target time sequence, constructing a plurality of isolation trees, selecting from the target time sequence a data point to be detected, traversing isolation trees among the plurality of isolation trees according to said data point, and, on the basis of the depths of said data point in the isolation trees, determining anomaly scores of said data point; on the basis of the anomaly scores of said point, determining an anomalous point list composed of anomalous points that have been detected; and, according to a time interval between the anomalous points in the anomalous point list, identifying an anomalous shaking interval. According to the present method, the problem of missed detection during anomaly detection on vehicle status information can be reduced.

The invention relates to the field of data processing, in particular to an operation state monitoring method and system for battery automatic short circuit antipole detection equipment, and the method comprises the steps: obtaining the current, voltage and temperature historical data of the battery detection equipment, calculating the data fluctuation degree through employing an HWKS algorithm, setting an initial k'value, and calculating the initial k 'value; clustering the historical operation data of different levels, analyzing the aggregation degree and local density of a cluster, correcting an initial k'value to obtain a corrected significant k value, taking the significant k value as a parameter value in an HWKS algorithm, obtaining the abnormal degree of the data of each level, performing weighted summation on the abnormal degree of the data of each level, and obtaining the abnormal degree of the data of each level. And the data processing module is used for acquiring the comprehensive abnormal degree of the data of each level, judging whether the data point is abnormal data based on the comprehensive abnormal degree, and monitoring the detection equipment. According to the method, the k value is adjusted in a self-adaptive mode, abnormal points in the data are accurately recognized, and therefore the accuracy of operation monitoring is improved.

The invention relates to the technical field of circuit board detection, in particular to a circuit board sensor nondestructive testing method based on big data, which comprises the following steps of: acquiring temperature, pressure, current and voltage data of key positions of a circuit board sensor, sending the data to a central database through real-time communication for summarizing, and synchronously performing data preliminary screening; and checking the integrity and availability of the data to form sensor summarized data. According to the method, the difference value between each data point and the adjacent data point is calculated, and the variation amplitude of the continuous data points is monitored, so that tiny abnormal fluctuation can be more sensitively captured, potential fault signals can be marked, an abnormal mode can be compared with a recorded fault case, and the fault mode and the frequent occurrence area of the fault mode can be identified; the fault mode is directly associated with the operation condition, the accuracy of fault prediction is enhanced, the dynamic adaptability of the detection system is ensured through monitoring adjustment of detailed fault analysis, and the efficiency of fault response is remarkably improved.

The invention relates to the technical field of data processing, and discloses a power grid fault early warning system based on machine learning and a related device, and the system is characterized in that an initial monitoring module identifies and identifies abnormal data points through monitoring the voltage, current and load of a power grid, and obtains an abnormal mode identification result; the load fluctuation analysis module positions the abnormal node according to the abnormal mode identification result to obtain abnormal node positioning information; the trend analysis and prediction module evaluates and predicts fault nodes and time according to the abnormal node positioning information to obtain a risk prediction result; the strategy formulation and response module formulates an operation instruction based on the risk prediction result to obtain a fault prevention measure; the fault simulation and test module performs virtual fault simulation by adopting fault prevention measures, and performs iterative optimization on a strategy to obtain a test evaluation result; and the fault early warning and optimization module optimizes a fault prevention strategy according to a test evaluation result, and a more accurate target fault early warning and a faster and more effective target response strategy can be obtained.

The embodiment of the invention provides a computer communication method and system based on the Internet of Things, and the method comprises the steps: constructing a multi-level system architecture, carrying out the preprocessing and feature extraction of an original data flow, recognizing the data characteristics through a time sequence analysis method, and constructing a dynamic data model; deploying a monitoring agent at a transmission node to collect network performance indexes in real time, and constructing a network quality evaluation model; for an incomplete sensor data flow, prior probability distribution is constructed based on a dynamic data model and a network quality grade, and an optimal estimation value of missing data is calculated by adopting a Bayesian reasoning framework and an iterative algorithm; establishing a mapping relation between a network state and an optimal parameter through reinforcement learning to realize self-adaptive adjustment and optimization; and grading the data according to reliability, extracting high-reliability data points as anchor points, designing an iterative refinement algorithm to realize information propagation, and fusing to obtain a complete sensor data stream. According to the method, the problems of poor data recovery accuracy, static parameter configuration and insufficient adaptability in a complex network environment are solved.

The invention relates to the field of foundation detection and multi-source big data intelligent analysis, and discloses a multi-source big data intelligent analysis fusion processing method based on foundation detection equipment, and the method comprises the following steps: carrying out the standardization processing, sampling frequency unification, missing value filling, and completion time and space alignment of multi-source data collected by the foundation detection equipment; based on the preprocessed multi-source data, constructing a similarity weight matrix between the data points for measuring the correlation between the data points; a manifold learning algorithm is adopted to embed multi-source data into a low-dimensional manifold space, and low-dimensional embedding representation used for describing a data local geometric structure is constructed; the method is based on a high-order Gaussian process model. By introducing the high-order Gaussian process method of multi-modal interaction modeling, the internal correlation of the modals and the complex interaction relationship between the modals can be comprehensively described. The problem that the nonlinear relation of the multi-modal data cannot be captured is solved, and more accurate feature extraction and fusion of the data are realized.

A method performed by a node for handling drift in data. The node obtains a dataset including a plurality of datapoints corresponding to a plurality of values of one or more dependent variables for a plurality of first features over a time period. The node determines, using machine learning and explainability, in the absence of determining whether or not the plurality of datapoints has a drift, whether or not there has been a change in respective one or more characteristics of a subset of the plurality of first features having a largest contribution to a variability of the datapoints in the plurality of datapoints based on a threshold from a first time period to a second time period. The node then initiates application of a drift policy on the plurality of datapoints based on a result of the determination.

The invention relates to the technical field of coal mine safety monitoring, and discloses a coal mine heterogeneous data visualization processing method combined with artificial intelligence, and the method comprises the steps: collecting multi-source heterogeneous data above and under a coal mine to construct a real-time heterogeneous data matrix; feature categories are divided based on data dynamic relevance, and stable data areas are screened through a dynamic analysis window to generate an optimization matrix; performing pattern recognition on the unstable region to extract an abnormal data cluster, and calculating an abnormal feature factor of a data point; analyzing the feature distribution difference to generate a stable feature index, and calculating an abnormal risk value in combination with the distance, the abnormal feature factor and the stable feature index; and mapping the center coordinate of the stable region to a three-dimensional space of the coal mine, and positioning a key information region in combination with the abnormal risk value. An apparatus includes a memory, a processor, and a computer program. The coal mine heterogeneous data processing efficiency and the safety monitoring accuracy are improved, intelligent positioning of the key information area is achieved, and powerful support is provided for coal mine safety production.

The invention discloses a data management method of a cross-domain data automatic clustering algorithm, which relates to the technical field of data processing, fully considers the characteristics of cross-domain data through distributed collaborative clustering, and adopts a clustering algorithm based on multi-domain kernel density estimation and adaptive neighbor search. According to the method, the bandwidth and the neighbor weight of kernel density estimation can be dynamically adjusted according to the distribution characteristics of data in different fields, the clustering structure in the data can be identified more accurately, and meanwhile, in the global clustering optimization process, a classification method based on fuzzy logic and evidence theory fusion is applied, so that the clustering optimization efficiency is improved. The feature weight of the field where the data points are located, the distance from the data points to the clustering center and global density information of clustering are comprehensively considered, it is ensured that clustering attribution of the data points is more reasonable and accurate, the problems of clustering errors and instability caused by the fact that a traditional clustering method cannot adapt to cross-field data complexity are effectively solved, and the clustering efficiency is improved. Therefore, a more reliable basis is provided for subsequent data analysis.

The invention relates to the technical field of industrial internet, and discloses a multi-dimensional data integration and dynamic risk assessment method for enterprise purchase anomaly detection, which comprises the following steps: establishing a reliability verification matrix of multi-dimensional purchase data, and extracting abnormal data points; identifying a cross correlation mode of the abnormal data points, and constructing a purchase behavior interaction graph of the target enterprise; analyzing a risk evolution path of the purchase risk factor, and calculating a risk conduction entropy value; and according to the risk conduction entropy value, generating a risk grading label of the purchase risk factor so as to output a purchase anomaly detection report of the target enterprise. According to the invention, the accuracy of enterprise purchase anomaly detection can be improved.

The invention discloses a hydraulic engineering supervision and analysis system and method based on digital twinning, and belongs to the technical field of hydraulic engineering operation monitoring and management, and the method comprises the steps: recognizing data points with periodic fluctuation characteristics and sparse spatial distribution, extracting a weak disturbance data subset, and building a response evolution curve; identifying a potential abnormal region through neighborhood statistics and clustering; for the abnormal region, inverting possibly existing micro-damage types and distribution directions based on historical time sequences and environmental boundary conditions to obtain an inversion defect feature set, and inputting the inversion defect feature set into a twinborn model to execute local heat-seepage-force coupling calculation to obtain a stress response evolution curve in a future preset period; calculating a risk score value according to the growth rate, step characteristics and frequency abrupt change of the curve, and outputting risk level and evolution trend information when the score exceeds a set threshold value; according to the invention, early recognition and dynamic evolution analysis of tiny hidden dangers can be realized, and the intelligent and refined level of hydraulic engineering safety supervision is improved.