142 results about "Data dimension" patented technology

The application is suitable for the technical field of financial early warning, and provides a financial early warning system, comprising a heterogeneous data fusion module, a dynamic knowledge graph construction module, a risk perception and prediction module and a self-adaptive early warning generation module. The system solves the technical problem that the existing financial early warning system has a single data dimension, is difficult to depict a risk correlation network, leads to one-sided early warning, has insufficient early warning capability for "contagious" crises caused by associated risks, and has a static and rigid early warning model, which lacks self-adaptation and evolution capability, leading to gradual degradation of long-term early warning performance, poor explainability of early warning results, limited decision support, and inability to develop accurate and effective risk mitigation measures, greatly weakening the actual decision support value of the early warning system. The application achieves the technical effects of accurate, explainable and self-adaptive early warning of associated and dynamic financial risks.

The application provides a data storage method, medium and computer program product for a large model. The data storage method for the large model comprises: in an inference stage of the large model, organizing quantized data to be stored and corresponding quantized parameters in a same storage tensor, wherein the storage tensor has multiple dimensions, and from an outermost layer to an innermost layer dimension, the dimensions comprise: a block dimension corresponding to a physical storage block in a paging management mechanism, each physical storage block having a continuous physical address range; an attention head dimension corresponding to an attention head defined for the large model; a word element dimension corresponding to a word element generated in the inference stage; a quantization grouping dimension corresponding to a quantization grouping of each word element on a feature dimension of each attention head; and a data dimension organizing quantized data and quantized parameters of each quantization grouping; and based on a logical storage structure of the storage tensor, storing the quantized data and the quantized parameters in a physical storage block of a cache.

This invention discloses a method and system for dynamic data credibility assessment that integrates multi-dimensional indicators, relating to the field of big data processing. The method includes: formulating multi-dimensional credibility assessment indicators based on business needs; acquiring raw data streams from power acquisition terminals and calculating real-time indicator values; based on an integrated assessment model, merging and calculating the real-time indicator values ​​and weights to output a comprehensive credibility score; triggering an early warning signal when the comprehensive credibility score is lower than a credibility threshold, and locating the specific indicator type and associated equipment causing the low credibility based on indicator contribution analysis; matching and executing a preset repair strategy based on the location results, and iteratively re-executing the assessment process after repair until the comprehensive credibility score meets the credibility threshold. This solves the problems of incomplete data dimension coverage and insufficient dynamic adaptability to massive multi-source data in existing power big data processing scenarios, which lead to one-sided credibility assessment results and data distortion.

The application relates to the technical field of intelligent manufacturing teaching and training equipment, and discloses a lean production training platform based on production big data analysis, which comprises an entity production line module, a multi-source heterogeneous data acquisition system, an edge computing and control hub and a lean production training terminal. The entity production line module is used for simulating a production operation process in a real discrete manufacturing environment. The multi-source heterogeneous data acquisition system is distributedly arranged on the entity production line module and is used for collecting multi-dimensional data in a production process in real time. Through cooperation of an industrial sensor array, a machine vision unit, an RFID (Radio Frequency Identification) unit and a data aggregation terminal, multi-dimensional data such as equipment states, process parameters and material tracking are collected in real time, millisecond-level perception of whole-factor data such as people, machines, materials, methods and environments in the production process is realized, and the problems of single data dimension and rough perception granularity of a traditional training platform are fundamentally solved, so that a rich and accurate data basis is provided for lean analysis.

This invention belongs to the field of power system technology and discloses a method and system for fault diagnosis and location of distribution lines. The method includes: synchronously collecting line operation signals from all nodes of a distribution line through a hierarchical distributed multi-source sensor network; generating a time-frequency feature matrix and a fusion feature set based on multi-source heterogeneous data preprocessing and line scenario adaptation; constructing a fault feature set by strengthening the weight ratio of different fault features using dynamic weighting factors; constructing a line diagnosis model adapted to the distribution line scenario based on the fault feature set to achieve line operation status identification and fault type classification; and constructing a multi-source information fusion location model based on improved particle swarm optimization to output the fault location result of the distribution line. This invention solves the problems of difficult spatiotemporal synchronization and limited data dimensions in traditional multi-source fusion by constructing a spatiotemporally aligned multi-source data preprocessing framework, achieving panoramic monitoring and providing more comprehensive data support for fault diagnosis.

This invention discloses an adaptive optimization method for SMT production line process parameters based on digital twins, belonging to the field of surface mount technology (SMT) data control technology. It includes constructing a digital twin to map SMT production line elements, reconstructing a model to calculate the SMT production line process status, deriving an SMT production line capacity optimization scheme, dynamically correcting process parameters, eliminating model overfitting faults caused by batch variations in incoming materials, and optimizing the joint simulation of equipment status and process parameters. This invention upgrades SMT production line process optimization from traditional reactive optimization to predictive optimization, reducing quality defects at the source and achieving a fundamental upgrade in the optimization model. Through a four-dimensional hierarchical digital twin and reduced-order model virtual sensing technology, it achieves real-time acquisition of key physical quantities that traditional methods cannot measure, such as thermal field distribution, stencil deformation, and dynamic gaps. This eliminates decision-making blind spots in process optimization, improves the accuracy and timeliness of optimization decisions, and thus fills the data dimension gap in process optimization.

This invention discloses an intelligent analysis method and storage medium for multi-population FATA axial spondyloarthritis based on adaptive dual-strategy pool optimization, comprising the following execution steps: Step 1, data acquisition; Step 2, data preprocessing; Step 3, parameter initialization and population; Step 4, construction of FATA update strategy pool and binary conversion strategy pool; Step 5, adaptive strategy iterative optimization; Step 6, outputting the optimal feature subset and using it for classifier diagnosis. This invention effectively expands the search space of high-dimensional heterogeneous axSpA data, improves the efficiency of obtaining the globally optimal feature subset, and reduces data dimensionality and computational complexity by using dual-strategy pool collaborative adaptation and multi-population parallel exploration, combined with an adaptive strategy learning mechanism, thus providing more accurate auxiliary diagnostic support for clinical practice.

The application provides a charging pile remote regulation and control method and system based on a cloud platform, and belongs to the technical field of new energy infrastructure and smart grid technology.The method comprises the following steps: collecting historical charging data of charging pile users and historical operation data of power grids in the regions where the users are located, generating user charging original data sets and power grid original operation data sets; and using a data mining algorithm to construct a user behavior and power grid state double-dimensional model and generate double-dimensional model data; through the construction of the user behavior and power grid state double-dimensional model, the user charging habits and the real-time operation characteristics of the power grid can be accurately captured, detailed and comprehensive data support is provided for the regulation and control of the charging pile, the problem of inaccurate regulation and control caused by single data dimension in the traditional mode is effectively avoided, and the scientificity and rationality of the charging regulation and control decision are improved.

The application relates to the technical field of model training, and discloses a lung cancer screening model training method and device based on gene transcriptome data, equipment and a medium. A basic model and reference transcriptome data are acquired. Core feature recognition is performed on the reference transcriptome data to perform data dimension reduction on the reference transcriptome data, and feature gene data is obtained. The basic model is pre-trained by using the feature gene data, and a preliminary training model is obtained. The preliminary training model is fine-tuned by using the reference transcriptome data, and an intermediate training model is obtained. In the case that the intermediate training model meets performance verification conditions, the intermediate training model is output as a lung cancer screening model. The beneficial effect is that core feature recognition and data dimension reduction are performed on existing reference transcriptome data, feature gene data suitable for large model training is obtained and used for training, the advance and objectivity of the gene transcriptome data are effectively utilized, and the screening timeliness and accuracy of the lung cancer screening model are improved.

The application discloses a wind turbine monitoring big data analysis system based on voiceprint recognition multi-mode fusion analysis and belongs to the technical field of power grid monitoring big data analysis; and is used for solving the technical problems of poor data processing and analysis effect and poor diversified early warning effect of the prior art; by integrating multi-mode data such as voiceprints, vibrations and temperatures, the traditional single-sensor monitoring data island phenomenon can be avoided, the data dimension can be improved from 3D to 23D, complete input is provided for fault diagnosis; the first identifier and the second identifier realize accurate association of data and equipment, and ensure the accuracy of subsequent preprocessing and analysis; environmental noise is removed through preprocessing, the signal-to-noise ratio of the voiceprint signal is effectively improved, and the feature extraction accuracy is improved, reliable data can be provided for subsequent fault analysis; based on label classification storage, the data security is improved; through the distributed storage characteristics of the block chain and the fault influence quantitative analysis, slight and serious fault grading early warning can be realized.

The present application relates to the technical field of data acquisition, in particular to an intelligent logistics distribution data acquisition intelligent control system and method, which constructs an acquisition task graph based on vehicle trajectory prediction, task target distribution and historical path mutation record, labels path sensitive section and replaceable node cluster, and generates an acquisition plan candidate set with path fault tolerance structure. Through the sensor scheduling engine, the sensor acquisition type, data dimension redundancy and energy sensitivity are comprehensively considered, the acquisition task is implemented cross-modal peak-shaving sorting and redundancy compression, a multi-modal acquisition strategy table is formed, and support nodes such as material transfer, signing confirmation and perishable goods state critical point are identified. When the vehicle approaches the support node, the acquisition preparation mechanism is triggered and data acquisition is performed, after the acquisition is completed, the distribution task target is checked, and according to the checking result, the path sensitive structure and the acquisition strategy are dynamically adjusted, realizing the intelligent closed-loop control of data acquisition in asynchronous environment.

The application discloses a kind of based on flow snapshot visual coding and timing behavior modeling computing power network DDoS attack anomaly detection method and system, first to original flow is carried out feature screening and normalization processing, and multidimensional flow feature is mapped into two-dimensional image representation, while retaining key space-time semantics Significantly reduce data dimension;Subsequently, a hybrid deep learning model is constructed, in which the flow snapshot visual coding feature extraction module based on EfficientNet efficiently captures the relevance between computing power task streams within a single time window, and its output is input into the LSTM module after structure remodeling as a time series to capture the flow timing evolution law caused by computing power scheduling;Finally, the spatiotemporal features are discriminated by a lightweight classifier after fusion.The present application fully meets the comprehensive needs of low resource overhead, high real-time and high accuracy of computing power network, and shows significant advantages in large-scale, high-bandwidth and distributed attack scenarios.

The application discloses a data processing method for a trusted data space, comprising: analyzing a data collaboration request by a trusted scheduling gateway, determining a temporary access token and local task data of an edge device, a data provider for collaborative processing of a multi-party collaborative task, a dynamic access route of the data provider, and a target data dimension required to be provided by the data provider; checking the access legality of the edge device based on the temporary access token and the local task data of the edge device; if the checking is passed, requesting the data provider to feed back target task data corresponding to the target data dimension based on the dynamic access route; desensitizing the target task data based on a data desensitization rule corresponding to the data provider to obtain desensitized task data; and sending the desensitized task data and the local task data of the edge device to a task execution party, so that the task execution party processes the multi-party collaborative task. The application guarantees data privacy and access security, and realizes efficient and trusted collaboration of multi-party data.

This application provides a network channel routing verification method and apparatus. The method constructs a directed graph based on the directed connection relationships of ports in the routing data, and uses graph algorithms to partition connected components and add audit rules based on data dimensions for verification. This allows for rapid and accurate detection of the connectivity and integrity of channel routes. During the verification process, it does not rely on the segment group order attribute in the routing data, avoiding shallow data quality issues that could prevent verification. Furthermore, it can verify the integrity of the channel carrying structure according to the business channel dimension, quickly aggregating all hierarchical channel routing anomaly audit results. For business channels with abnormal aggregated audit results, a secondary verification is performed to ensure the accuracy and reliability of the verification results. This helps vendors rectify data according to the business channel dimension and achieve rapid results.

A system and method for using enhanced QR codes in a call to action, that provides enhanced functionality for generating a call to action element or providing personalized content when scanned, and that combine additional data dimensions with existing QR code technologies to expand the QR code capability beyond what is enabled by standard QR code specifications.

The application discloses a data layered transmission method and system of an intelligent fusion terminal, and relates to the technical field of digital information transmission.The method comprises the following steps: obtaining historical multi-dimensional working condition data and preprocessing the same to construct a historical state feature vector set; calculating the dynamic weight of each data dimension based on the historical state feature vector set; clustering the historical state feature vector set and weighting the calculation of the Euclidean distance in the clustering algorithm based on the dynamic weight to generate a plurality of transmission state clustering clusters; introducing a profile coefficient to construct a clustering quality dynamic evaluation and optimization mechanism and determining whether to update the transmission state clustering cluster in real time; constructing a dynamic reward function to improve a Q-Learning algorithm and generating a state-action Q table based on the transmission state clustering cluster; and executing a transmission strategy suitable for the current transmission event working condition state based on the state-action Q table.The application can effectively balance the local action optimization and the global resource efficiency and improve the utilization rate of communication resources.

The application discloses a gene query interpretation and health management interaction method and system based on a big data model, wherein the method comprises the following steps: acquiring multi-modal health data of a user, wherein the multi-modal health data comprises gene data, phenotype data, living habit data and sociological environment data, and each data dimension is acquired through a standardized collection process; performing cleaning, normalization and correlation integration on the multi-modal health data, and constructing a unified health portrait of the user, wherein the unified health portrait comprises genetic characteristics, physiological phenotype characteristics, living behavior characteristics and social environment characteristics; calling an intelligent agent processing engine to perform deep analysis and processing on the unified health portrait; and generating a personalized health management scheme based on a gene interpretation conclusion and health management suggestions, wherein the scheme can be dynamically adjusted according to the health state change of the user, and a scientific, comprehensive and personalized solution is provided for individual health management.

The present application relates to a kind of 3D twin intelligence examination device and system to stab training, belong to intelligent training examination technical field, including inductive protective clothing, no wearing motion capture terminal, smart bracelet, multi-source data synchronization module, intelligent simulator, triplex screen host, the inductive protective clothing includes multiple combined armors and abdominal armor, multiple the combined armors are equipped with protective bending part and fixing belt part, several the fixing belt part is formed by the fastening structure of elastic band and velcro combination, multiple protective bending part includes fixed bending plate, elastic protective pad and conversion convex shell, the present application solves the problem of single data dimension in traditional assassination training examination, big artificial error in judgment, no tactical analysis, training injury prevention and control loss, AI guidance blank and poor review effect, realize the digitization, standardization, intelligentization of assassination training examination, while guaranteeing data local processing, satisfy information security requirement.

The present application belongs to the technical field of medical health, and particularly relates to a youth physique dynamic monitoring and intervention method and system based on multi-modal fusion. Physique data is collected and processed through a stereoscopic vision sensor array and a deep learning algorithm, multi-modal psychological data is collected through an AI drawing interactive process, and a personalized intervention scheme is generated through cross-modal semantic fusion and intelligent decision-making, thereby solving the problems of single data dimension, imperfect evaluation system, homogeneous intervention and scattered data of a traditional monitoring system, and realizing full-process intelligent and personalized dynamic management of youth physique health.

This application relates to the field of data management technology and discloses a data generation method, apparatus, device, and storage medium based on a data warehouse. The method includes: acquiring an original data table from a data warehouse and determining atomic indicators based on the original data table; acquiring target indicators and associated data dimensions of the atomic indicators under target constraints; determining indicator fields and dimension fields based on the target indicators and the data dimensions; summarizing the original data table based on the indicator fields and the dimension fields to obtain a summarized data table; and aggregating the summarized data table based on the dimension fields to generate a target data table. This application can effectively improve data processing efficiency and flexibility, achieve automated and efficient data development, and reduce data development costs.

The application discloses a complex equipment-oriented artificial intelligence high-quality data set construction method, equipment and medium, the data set construction method firstly carries out preprocessing to the obtained multi-source heterogeneous original data, then obtains the uncertainty score of each data sample through uncertainty quantification, and then combines the uncertainty score and multiple data dimensions to construct a comprehensive evaluation index to determine the sample level quality score; in data enhancement, data enhancement conforming to the physical knowledge constraint is carried out on the high uncertainty, low quality score and sample sparse area to obtain a synthetic sample; then, after feature extraction and alignment of the original sample and the synthetic sample, a dynamic fusion strategy is executed to obtain a high-quality data set. The high-quality data set is helpful to improve the model robustness and generalization ability. The application further constructs a closed-loop feedback mechanism for high-quality data set application, realizes the association and coupling of data set construction and model demand, and enables data-model to evolve cooperatively.

This paper relates to a generator set control method and device based on multi-timescale features. The higher-level scale provides guidance information to the lower-level scale; the weekly plan provides the unit start-up and shutdown status for the day-ahead plan, and the day-ahead plan provides the basic output points of the units for the intraday plan. At the weekly scale, the constraints of unit timing coupling can be ignored, and a dataset is generated by feeding back from the lower-level scale to the higher-level scale. This results in smaller data dimensionality and lower computational cost for data generation. Then, based on a neural network, the start-up and shutdown status of the units is predicted, and reliable start-up and shutdown statuses are selected and passed down, eliminating the need for feasibility remedial measures for start-up and shutdown statuses and reducing computational complexity. At the day-ahead and intraday scales, the start-up and shutdown statuses of the units determined by the higher-level time scale are used to identify more redundant constraints and accelerate the solution speed. This paper utilizes the features of different time scales to solve the unit combination problem, which can accelerate the solution speed while ensuring solution quality, and the solution speed is more stable.

This disclosure relates to the field of computer technology, specifically a system monitoring method and apparatus, a computer program product, and an electronic device. The system monitoring method includes: constructing a background information dataset based on historical system indicator data, wherein each piece of background information data in the dataset includes at least an application dimension, an interface dimension, and a data dimension; extracting semantic description information for each dimension of each piece of background information data, and constructing an initial background template based on the semantic description information corresponding to each piece of background information data; constructing a reference background template based on system indicator data within a target time period based on a preset time interval, and performing semantic matching between the reference background template and the initial background template to monitor system operation based on the matching results. This disclosure can improve the accuracy of system monitoring and enhance the efficiency of system monitoring and maintenance.

This invention discloses a method and system for semantic entropy feature evolution and register security processing. The security processing method includes: real-time monitoring of semantic fluctuation feature indicators of asynchronous communication stream sequences; constructing and generating a time-enhanced dynamic semantic vector containing historical context backtracking features; mapping the arbitration result to the dimension index corresponding to the physical execution layer to generate a binary dynamic masking operator; obtaining the pointer to the underlying hardware direct addressing memory address mapped by the privileged addressing interface to obtain a de-identified situational awareness feature vector set. By strictly aligning the dynamic masking operator and vector synthesis computation in time, the accompanying physical annihilation of sensitive data dimensions is achieved at the register output level, realizing computation-as-isolation and fundamentally eliminating the microsecond / nanosecond-level security vacuum caused by software audit lag.

The application discloses a Parkinson disease severity recognition method based on tensor ring decomposition and region segmentation, relates to the technical field of machine learning recognition, and comprises the following steps: collecting VGRF signals, accurately dividing the VGRF signals according to target personnel indexes, gait window indexes, foot indexes, set region indexes and sampling point time indexes, constructing a five-order time domain tensor, and improving the learning ability for large-dimension tensors; converting the five-order time domain tensor into a frequency domain five-order tensor, extracting low-rank structures in the frequency domain five-order tensor as core tensors, reducing the data dimension of calculation, and improving the learning power of a classification model; extracting statistical features of patients in gait, feet and each set region in the core tensors, further maintaining the correlation features between each dimension of data on the basis of reducing the data dimension of calculation, and then accurately distinguishing the severity of Parkinson disease of the patients and improving the accuracy of the classification model in recognizing the disease severity of the patients.

The application relates to the technical field of network security, in particular to an AI-based multi-dimensional network attack tracing and early warning system, which comprises a network security management center, a multi-source data acquisition module, an AI multi-dimensional analysis module, a defense matching module and an early warning operation module, the AI multi-dimensional analysis module comprises an attack identification submodule, a path restoration submodule and a source positioning submodule; the application collects multi-source data through fusion, avoids feature missing caused by single data dimension, reduces the missing judgment probability, simultaneously adopts a fusion model to identify attack types and intensity, improves attack identification accuracy, constructs an attack propagation graph and accurately locates an attack source, solves the problems of low efficiency and poor accuracy of traditional tracing, simultaneously generates a graded early warning based on a risk level, ensures that attacks of different severity levels are responded to correspondingly, realizes automatic matching of defense strategies, avoids protection delay, and improves the active defense capability of network security.