85 results about "Data level" patented technology

Aiming at privacy protection mean value estimation of multi-dimensional numeric data, a personalized privacy protection mechanism meeting-localization differential privacy is designed, and personalized privacy protection of a user level and a data level is provided. According to the invention, each user can select one privacy protection level from a plurality of preset privacy protection levels according to own privacy protection requirements, so that personalized privacy protection of the user level is realized, which is the first personalized privacy protection. And based on the expectation of the minimum estimation variance, determining an optimal dimension extraction parameter, and randomly extracting part of dimension data from all dimensions to carry out disturbance submission. And a user scoring strategy is adopted to determine a distribution strategy of privacy protection parameters, so that personalized privacy protection on a user data level is realized, which is second personalized privacy protection. And finally, a weighting factor is constructed based on the expectation of the estimation variance, and a weighted combination mode is adopted for mean value estimation under multiple privacy levels, so that the accuracy of the overall mean value estimation is further improved.

The invention relates to the technical field of production quality control, and discloses a production data hierarchical processing and quality tracing system. The system comprises a production data acquisition module, a data hierarchical processing module, a quality fluctuation index module, a multi-dimensional quality difference analysis module and an abnormal source positioning module. The production data acquisition module acquires a real-time parameter data sequence of each production link; the data layering processing module is used for layering the data into a plurality of data hierarchies; the quality fluctuation index module obtains a quality fluctuation index based on the fluctuation condition of the data hierarchy; the multi-dimensional quality difference analysis module analyzes the multi-dimensional quality difference based on the index; and the abnormal source positioning module positions a production abnormal source based on the multi-dimensional quality difference. According to the system, through hierarchical data processing, quality fluctuation quantification and multi-dimensional difference analysis, accurate positioning of a production abnormal source is realized, and systematicness and accuracy of production data processing and quality tracing are improved.

The invention relates to the technical field of multi-source data fusion, in particular to a multi-source data fusion-based intelligent detection method for a transportation state of combined transportation of iron and water, which comprises the following steps of: in a transportation process of combined transportation of iron and water, obtaining at least two monitoring data sequences, and in a process of carrying out multi-source data fusion on all the monitoring data sequences, carrying out multi-source data fusion on all the monitoring data sequences; performing feature extraction on each monitoring data sequence through data level fusion and feature level fusion to obtain at least two feature sequences; classifying all the feature sequences, obtaining an optimal attention mechanism according to the category to which each feature sequence belongs, performing decision-level fusion on all the feature sequences according to the optimal attention mechanism of each feature sequence to obtain a transportation state detection result of the combined transportation of the iron and the water, and obtaining a detection result of the transportation state of the combined transportation of the iron and the water through analysis of a self-adaptive attention mechanism. And the robustness of transportation state detection based on a multi-source data fusion technology in a railway-iron combined transportation scene is improved.

The invention discloses a segmented multi-source shoreline extraction and calibration method and system based on physical prior, and particularly relates to the technical field of remote sensing image processing and computer vision, a consistency evaluation and reliability measurement mechanism of a multi-source surface water product is integrated on a data level, and a high-quality training sample set is constructed by using water persistence characteristics; on the model level, a feature extraction framework fusing physical prior guidance and a geometric self-adaptive operator is researched and developed, a multi-dimensional physical constraint and noise tolerance loss function is constructed, and pixel-level accurate recognition and topological rigorous quality closed-loop calibration of a shoreline under a large-range remote sensing image are achieved.

The invention discloses a sensitivity data acquisition method and system based on environmental factors, and the method comprises the steps: obtaining a data set in a preset time period, the data set comprising meteorological data, environmental protection data, pollen concentration data and doctor seeing data; performing model training by using the data set to construct a prediction regression model; calculating each index coefficient and intercept based on the prediction regression model, wherein each index coefficient at least comprises a relative humidity coefficient, an air pressure coefficient, a temperature coefficient, a pollutant concentration coefficient, a pollen concentration coefficient, a precipitation coefficient and a wind speed coefficient; obtaining current environment data and future predicted environment data, wherein the current environment data and the future predicted environment data are in one-to-one correspondence with the index coefficients; and calculating the sensitivity data based on the current environmental data, the future predicted environmental data, the index coefficients and the intercept, so that the sensitivity data can be graded, and protection early warning can be provided for users in time according to the grading condition.

The invention discloses an ecological protection and restoration ecological benefit assessment method, which comprises the following steps: S1, N-level division of assessment indexes: dividing first-level indexes, and dividing second-level to N-level indexes in a tree structure based on the first-level indexes; s2, index data calculation: calculating and obtaining N-level index data, and then calculating and summarizing data level by level until first-level index data is obtained; and S3, metric value calculation: determining a weight coefficient of each first-level index by a structure entropy weight method, and calculating the first-level index data by a weighted summary algorithm to obtain an ecological protection and restoration ecological benefit metric value. According to the scheme, the structure entropy weight method and the weighted summary algorithm are combined, the uncertain influence possibly existing due to the difference of experts on index understanding can be reduced, the accuracy of weight coefficient determination is improved, and then the evaluation accuracy of the ecological protection and restoration ecological benefit measurement value is improved.

The invention discloses an electric vehicle driving data management system, which comprises a multi-source data acquisition module, a processing module, a grading judgment module, a dynamic fusion module, an improvement module and an interaction feedback module, the multi-source data acquisition module is used for acquiring various driving data of the electric vehicle in real time; the processing module is used for receiving and processing various mass driving data; and the grading judgment module is used for dynamically judging the data grade and executing a grading storage strategy. The invention belongs to the technical field of electric vehicle driving data management, and aims at solving the problems that in the prior art, a data storage mode is single, distinguishing is not carried out according to the importance of driving data, and meanwhile a large deviation exists between a prediction result and the actual endurance mileage. The method has the technical effects that the data can be conveniently divided into the high importance level and the low importance level, the differentiated storage strategy is executed, and meanwhile, the actual energy consumption condition of the vehicle can be reflected more accurately.

The invention provides an industrial process zero sample fault diagnosis method based on a similarity consistency relation, and relates to the technical field of fault diagnosis and classification. From the perspective of similarity between faults, on one hand, understanding of experts on the faults is deeply mined, and the similarity between fault categories is constructed based on the knowledge level; and on the other hand, the model is trained by using the known fault data, the data features of the known faults are extracted on the data level, and the similarity between the fault features is quantified from multiple angles. By solving a group of consistency parameters, mapping of faults from knowledge similarity to data similarity is constructed. Through the mapping, the abstract and comprehensive similarity relation of the knowledge level can be specifically explained by utilizing the similarity measurement of the data level, and the diagnosis of unseen class faults can be effectively supported.

The invention discloses a multivariable time series data prediction method based on depth sub-feature decomposition. The method comprises the following steps: firstly, recording and collecting time series data; and secondly, constructing a multivariable time series data prediction model, inputting the collected time series data, and carrying out characteristic decomposition to obtain a prediction result. And finally, training and testing the multivariable time series data prediction model. According to the method, the significance of the key time variable on the data level is improved, decoupling of complex mixed time sequence data is achieved, the time variable capturing difficulty is reduced, and the prediction precision and robustness of the model are improved.

The invention discloses a big data problem clue mining method based on a mutual exclusion rule. Relates to the technical field of big data supervision, in particular to a big data problem clue mining method based on a mutual exclusion rule. According to the method, 'exclusive contradiction between data records' is taken as a core entry point and analysis dimension, and key features which violate logic consistency are accurately captured on a data level. The method comprises the following steps: acquiring a multi-source heterogeneous data set, and preprocessing the multi-source heterogeneous data set to obtain an initial feature set and an entity subgraph; extracting mutually exclusive line features according to the initial feature set; based on the mutually exclusive line features, evidence features are generated; constructing a dual-channel model to calculate a mutual exclusion risk; using a decision tree improved algorithm as a rule enhanced feature channel; and constructing a graph relation perception channel: aggregating the abnormal probability of the feature space and the abnormal probability based on the graph structure to obtain a risk score through a dynamic weighting mode.

The invention discloses an auxiliary driving vehicle accident evidence chain tracing method and system based on a fault tree. The method comprises the following steps: acquiring auxiliary driving function state data, vehicle operation data and time sequence data of a driver state in a certain time period before an accident occurs; analyzing the time series data, and obtaining an abnormal data item list according to a predetermined threshold value; the abnormal data item list is input into a pre-constructed fault tree, and events are connected in series layer by layer through AND gates and OR gates to form a plurality of accident chains; judging an accident chain corresponding to the abnormal data item list according to the logic of the fault tree, finally determining a basic event of the accident chain, and judging an error passing party according to the basic event; and recording a judgment path of the fault tree, and generating a fault analysis report. According to the invention, accurate traceability can be realized from the data level, and the systematicness and accuracy of the accident evidence chain are greatly improved.

The invention discloses an implicit image degradation estimation method based on decoupling contrast learning, and belongs to the technical field of image blind super-resolution, and the method comprises the steps: sampling image blocks in different LR images with the same degradation information to construct a positive sample set; on the data level, it is guaranteed that task irrelevant information such as content and texture does not affect degradation feature learning, and meanwhile data amplification is achieved by continuously and circularly pairing sampling image blocks; in the training process, a feature enhancement strategy is introduced to increase the number of positive samples of the same degradation type, so that the sample diversity of each degradation type is further expanded in the training process to ensure that the network more fully learns degradation information. The implicit image degradation estimator trained through the method is applied to a blind image super-resolution network based on implicit degradation feature guidance, the degradation information modeling quality of an LR image can be effectively improved, and then the reconstruction effect from the LR image to an HR image is improved.

The invention provides a method and a system for hierarchical storage of block chain data in a city update field. Wherein the project data of the city update project is divided into a plurality of data levels according to the life cycle stage of the city update project, the project data of each data level is classified and stored according to the update frequency and the access demand, and the project data with high-frequency update and high access demand are stored on the chain of the block chain. Project data with low-frequency updating and low access requirements are stored under the block chain; a multi-level index structure is constructed on the chain, and the multi-level index structure is used for quickly positioning the project data stored under the chain and ensuring the integrity and traceability of the data on the chain at the same time; a data migration strategy is set, and when it is detected that the access mode of the item data in any data level changes, item data migration between the data levels is triggered. According to the technical scheme provided by the invention, the efficiency and safety of urban update project management are improved.

The invention discloses a deep learning-based cross-department government affair data security sharing and analysis method, and relates to the technical field of data sharing and analysis. Comprising the following steps: acquiring multi-source data across departments, converting the multi-source data into standard data through data cleaning verification, full-amount acquisition implementation and data type unified conversion processing in sequence, and performing distributed storage; constructing a knowledge graph, performing semantic analysis on the data stored in a distributed manner, then analyzing the data, calculating data levels, and constructing a multi-level retrieval directory corresponding to the data levels; the technical key points are as follows: hidden association between entities is mined through a knowledge graph, deep learning is performed to perceive risks in real time, and compared with a passive storage mode in the prior art, upgrading of data from only sharing to safe use is realized, so that sharing efficiency can be guaranteed, and the depth and security of data analysis can be improved; the method meets the requirements of cross-department collaborative decision making, is good in use effect, and has a good use prospect.

A method for training a machine learning algorithm including uncertainties. The method includes the following steps: for each point in time of the plurality of points in time, determining in each case an influence, which the data detected at the corresponding point in time have on uncertainties instantaneously contained in the initial model, for each point in time of the plurality of points in time, determining a resolution of the corresponding detected data based on an established data level and on the respective influence, which the corresponding data have on uncertainties instantaneously contained in the initial model, for each point in time of the plurality of points in time, transferring the detected data to the control unit based on the corresponding determined resolution, and retraining of the initial model by the control unit based on the data transferred to the control unit.

The present invention discloses an implicit image degradation estimation method based on decoupled contrast learning, which belongs to the field of image blind super-resolution technology. The method samples blocks from different LR images with the same degradation information to construct a positive sample set, which ensures that task-irrelevant information such as content and texture will not affect degradation feature learning at the data level. At the same time, data amplification is achieved by continuously cyclically pairing sampled blocks. In the training process, a feature enhancement strategy is introduced to increase the number of positive samples of the same degradation type, thereby further expanding the sample diversity of each degradation category during the training process to ensure that the network learns degradation information more fully. Applying the implicit image degradation estimator trained by the present invention to a blind image super-resolution network guided by implicit degradation features can effectively improve the degradation information modeling quality of LR images, thereby improving the reconstruction effect of LR images to HR images.

Disclosed is a method for managing gas data, which is implemented by at least one processor of an Internet of Things (IoT) system for managing the gas data. The method comprises: obtaining to-be-stored-gas data and downstream user features; determining a user importance level based on the downstream user features; determining accessing frequency distribution features of the to-be-stored-gas data; determining a risk degree of data through a second prediction model; constructing query feature vectors based on pipeline data, and determining a risk degree of the gas pipeline based on the query feature vectors; determining a gas data level; determining a data redundancy level; generating redundant data blocks of the to-be-stored-gas data; and storing the to-be-stored gas data and the redundant data blocks in the at least one storage node of the at least one sub-data center based on loading distribution features.

This application provides a method for selecting training corpus based on the results of a text classification model fusion using multiple loss functions. At the model level, by fusing multiple loss functions, the text classification model adaptively learns the weights of each loss function on the classification performance, thereby improving its robustness. At the data level, based on the results of the text classification model fusion using the aforementioned multiple loss functions, variance calculation is used to determine the quality of the output category division of the training sample data, identifying poor-quality data for review and processing. The text classification model is then retrained based on the processing results to improve its classification or prediction performance. By calculating the confusion between output categories, a quantitative score is generated for the classification system of the text classification model, serving as the basis for adjusting the classification definitions within the model, thereby improving its prediction performance.

This application discloses a data grading method for the evolution of data space information, including: in response to the process of generating second data based on a data operation on first data, obtaining the first data level of the first data and the operation type of the data operation; determining the correlation strength evaluation value between the first data and the second data based on the first data and the second data, as well as the operation type of the data operation; and determining the second data level of the generated second data based on the first data level, the correlation strength evaluation value, and the operation type of the data operation. This method solves the problem of repeated grading overhead in the dynamic evolution of data and successfully builds a traceability system for data evolution levels, providing technical support for the efficient management and intelligentization of data space.

The invention relates to the technical field of power system reliability evaluation, in particular to an offshore wind power submarine cable transmission power probability safety evaluation method and system. According to the method, multi-dimensional data management, dynamic prediction modeling and self-adaptive decision optimization are creatively fused, so that double breakthrough of safety assessment precision and robustness is realized. Abnormal data removal, blank value filling and normalization are carried out on the data in the data level, data noise is reduced, and key features are completely improved; in a prediction level, an Att-BiLSTM model is constructed, and wind power output prediction is realized through bidirectional time sequence feature extraction and attention weight dynamic distribution; in the aspect of probability safety assessment, a deep reinforcement learning framework based on a PPO algorithm is designed, the submarine cable power safety margin is dynamically described, full-chain technical support is provided for offshore wind power high-reliability grid connection and power transmission safety monitoring, and the method has wide industrial application prospects.

The application discloses a kind of track traffic equipment health degree prediction method and system based on multi-source information fusion, it is related to track traffic equipment health degree prediction technical field, including: the multidimensional fault index data of track traffic equipment is collected;Data dimension reduction processing is carried out;The time series analysis is carried out to the fault feature vector after dimension reduction processing in combination with time stamp data;Health degree prediction is carried out by integrating algorithm model coupling time series analysis result and multi-source data;Based on health degree prediction value, the overall operation state of equipment and potential failure risk are evaluated and early warning.The application realizes the cooperation of data level, algorithm level and optimization level, the fault feature output by PCA model is one of the inputs of Prophet and XGBoost model, and XGBoost also combines other multi-source data for comprehensive prediction, the accuracy, real-time performance and reliability of prediction are ensured through effective model cooperation, which helps to discover potential faults and problems in time.

Real-time monitoring and analysis of data anomalies is described. An example system for detecting data anomalies includes a network interface configured to receive individual information from each of a plurality of individual devices, the individual devices being computing devices. The system also includes a processing unit configured to extract relevant data indicators from the individual information using predefined algorithms, compute a statistical value indicative of a collective data level of the plurality of individual devices by integrating the relevant data indicators, and dynamically adjust data monitoring parameters based on real-time data to enhance accuracy. The system also includes a memory unit configured to store the individual information, the relevant data indicators, and the computed statistical value. The system also includes a feedback module configured to provide personalized data management recommendations associated with the individual devices based on their collective data levels and predefined data relief protocols.

The application provides a sintering flue gas CO emission concentration prediction method, and relates to the field of sintering process pollutant treatment. It comprises the following steps: S1, data acquisition and preliminary rejection; S2, constructing a sample data set and preprocessing it; S3, characteristic value screening, based on the preprocessed sample data set, combining a nonlinear correlation analysis method and a model-driven recursive elimination strategy, constructing an optimal feature subset; S4, integrated modeling, adopting a Stacking integrated learning framework, integrating the advantages of XGBoost, random forest and gradient boosting tree algorithms, and constructing an effluent CO concentration prediction model. Through multi-layer screening of characteristic values, the application clearly determines the dominant action mechanism of each process link on CO generation from the data level, and significantly improves the prediction accuracy and robustness compared with existing models.

The application discloses a segmented multi-source shoreline extraction calibration method and system based on physical prior, and particularly relates to the technical field of remote sensing image processing and computer vision, and integrates a consistency evaluation and reliability measurement mechanism of multi-source surface water products at a data level, and constructs a high-quality training sample set by using water body persistence characteristics; at a model level, a feature extraction architecture is developed by fusing physical prior guidance and a geometric self-adaptive operator, a multi-dimensional physical constraint and a noise tolerance loss function are constructed, and pixel-level accurate identification of a shoreline under a large-range remote sensing image and quality closed-loop calibration with a rigorous topology are realized.

The invention relates to a rapid identification and analysis method for discrete points of continuous data. The method comprises the following steps: step 1, defining basic information; 2, calibrating the data level of the normal data value in the initial operation period through the data dispersion; 3, identifying, tracking and locking discrete points; the method has the advantages that the requirement for data processing computing power is lowered, and the effectiveness and timeliness of discrete point grabbing are improved.

The application discloses an engineering project cost whole-process collaborative management method and system, relates to the technical field of engineering project cost management, and comprises a cost matter main line construction module, which is used for receiving project estimation, design budget estimate, bidding control price, contract price, process measurement, change visa, claim and item data of final account, performing unified matter code mapping and cross-stage integration processing, and generating cost matter unit data. In the application, the project estimation, the design budget estimate, the bidding control price, the contract price, the process measurement, the change visa, the claim and the item data of the final account are subjected to the unified matter code mapping and the cross-stage integration processing by the cost matter main line construction module, the cost matter unit data is generated, cross-stage item integration record data is formed by the different-stage item data, and thus a stable main line structure is established at the data level, and consistent organization and centralized management of the cross-stage item data are realized.

The invention discloses a target detection method and device based on multi-sensor fusion and an automatic driving vehicle, and relates to the field of target detection, and the method comprises the steps: obtaining image data and point cloud data after completing the time synchronization of an image collection module and a radar module; determining first position information and categories of the S first targets under the image coordinate system according to the image data and a preset deep learning algorithm; determining second position information of T second targets in the point cloud coordinate system according to the point cloud data and a preset semantic segmentation algorithm, and further determining spatial perception information; and completing association of the first target and the second target belonging to the same detection target based on the conversion matrix. According to the scheme, the first target is determined by directly utilizing the image data, the second target is determined by utilizing the point cloud data, and fusion on the data level is realized based on the conversion matrix, so that respective advantages of the image acquisition module and the radar module are well utilized, and the perception precision of the target is improved.

This invention discloses a method for inverting and cascading early warning of hidden dangers in reservoir groups based on spatiotemporal network coupling. The method includes: acquiring hydrological time-series monitoring data and geospatial attribute data of reservoir groups in the region, and constructing a multi-source feature matrix including water level data and elevation information; filtering water level data, constructing a cumulative deviation curve, and judging the hidden leakage trend; constructing a multi-dimensional resilience assessment model based on water level data, obtaining the comprehensive resilience score of each reservoir node, and mapping the comprehensive resilience score to the node's anti-disturbance ability. By fully exploring the value of existing water level and rainfall monitoring data, and using neighborhood dynamic benchmarks and residual accumulation models, natural evaporation and meteorological interference can be effectively filtered out at the data level, and hidden leakage signals inside the dam can be interpreted from conventional hydrological data, providing a low-cost and universally applicable safety monitoring path for small reservoir groups.

The present invention discloses a method for controllably generating bolt defects guided by perspective and attributes. A generative adversarial network is constructed, and the generator of FastGAN is selected as the basic model for the generation part. Perspective and attribute information are introduced into the generator as conditions to guide the generation, and a residual cross-layer excitation ResSLE is proposed to strengthen the gradient information flow and increase the controllability of generation. In the discriminant part, a U-Net self-supervised reconstruction discriminator USRD is proposed, and a unique local cropping method Shape-crop is designed for T-shaped bolts, thereby enhancing the ability to generate bolt texture details and extract features. This bolt defect generation method can controllably generate bolt defect images with specific perspectives and attributes under the condition of a small number of samples, and assist in improving the accuracy of the bolt defect recognition network at the data level.

A big data problem clue mining method based on mutual exclusivity rules. It relates to the technical field of big data supervision, and particularly relates to the technical field of a big data problem clue mining method based on mutual exclusivity rules. The present application takes the mutual exclusivity contradiction between data records as the core breakthrough point and analysis dimension, accurately captures at the data level, and extracts the key features that violate logical consistency. The method comprises the following steps: obtaining a multi-source heterogeneous data set and pre-processing to obtain an initial feature set and an entity subgraph; extracting mutual exclusion features according to the initial feature set; generating evidence features based on the mutual exclusion features; constructing a double-channel model to calculate the mutual exclusion risk: using a decision tree improvement algorithm as a rule-enhanced feature channel; constructing a graph relationship perception channel: through a dynamic weighting method, the abnormal probability of the feature space and the abnormal probability based on the graph structure are aggregated into a risk score.