25 results about "Synthetic data sets" patented technology

The application discloses an urban waterlogging intelligent prediction method and system based on a physically constrained enhanced video generative adversarial network. The method comprises the following steps: constructing a meteorological geographic data set containing static terrain data and dynamic rainfall data; generating waterlogging simulation data based on a physical numerical simulation model, and matching the waterlogging simulation data with the meteorological geographic data set into a comprehensive data set; constructing a physically constrained enhanced video generative adversarial network model, wherein the model comprises a generator and a discriminator; constructing a multi-angle loss function system that fuses an adversarial loss, a time sequence consistency loss, a data supervision loss and a water quantity balance constraint physical loss; and alternately optimizing the generator and the discriminator until the model converges, and outputting a spatiotemporally continuous waterlogging inundation depth prediction result. The application explicitly embeds a physical hydrological law into a deep learning network, and eliminates the physical logic contradiction of a pure data-driven model in continuous spatiotemporal sequence prediction through water quantity balance constraint.

The disclosure provides a data set distillation method based on a boundary perception diffusion model, comprising: pre-training an initial expert model using an original training set, fitting a decision boundary of the initial expert model to a true boundary by adjusting a weight parameter; generating a multi-condition fine-tuning loss reflecting different categories using a class probability predicted by the trained expert model as a confidence weight, and guiding fine-tuning of the diffusion model through minimization of a weighted sum of the multi-condition fine-tuning loss; performing multi-condition sampling using the fine-tuned diffusion model, and guiding the image generation of the diffusion model to approach the decision boundary region between the two categories by introducing labels to weight and average the noise prediction results of the target class and adjacent competing classes according to a mixing coefficient; and dynamically mixing boundary discrimination samples generated through multi-condition sampling and intra-class representative samples generated through single-condition sampling according to a preset ratio to construct a lightweight synthetic data set as a final result of data set distillation.

Systems and methods include determination of a plurality of event types, determination of a plurality of associated metrics for each of the plurality of event types, generation of synthetic time-series data of each of the metrics associated with the event types for each of the plurality of event type, the synthetic time-series data representing metric anomalies and events of the event type, training of a first system to generate anomaly values based on the synthetic time-series data, training of a second system to classify event types based on the anomaly values, determination of an anomaly value for each data instance of actual time-series data, determination of event types for each determined value, selection of data instances based on the determined values and event types, reception of an event type for each selected data instance, and re-training of the second system based on the selected data instances and event types.

The application discloses a method for obtaining soil strength indexes based on soil layer geophysical prospecting data, and comprises the following steps: S1, obtaining the test depth and equivalent shear wave velocity of different measuring points in a target site; obtaining the drilling samples of each measuring point at the corresponding depth, obtaining the corresponding internal friction angle and soil type; preprocessing the collected data to obtain a comprehensive data set; S2, selecting the test depth and equivalent shear wave velocity in the comprehensive data set as basic features, constructing derived features, performing feature standardization, and obtaining a data set; dividing the data set into a training set and a test set; S3, based on the training set obtained in S2, optimizing and training the soil classification model and the internal friction angle prediction model; S4, obtaining the soil strength index internal friction angle based on the soil classification model and the internal friction angle prediction model obtained by the training in S3. The method of the application realizes high-precision prediction of soil strength indexes by constructing derived features and introducing physical range constraints.

The application provides a data processing method and device, equipment, medium and product, relates to the technical field of data processing, and comprises the following steps: a first data set with a complete inverted index structure is established according to heterogeneous data provided by different data sources; a comprehensive data set is generated according to data in the first data set and statistical characteristics thereof; data consistency indexes, data integrity indexes and data reliability indexes of each data source are calculated according to the comprehensive data set; data related to a query statement input by a user is searched according to the comprehensive data set, the data consistency indexes, the data integrity indexes and the data reliability indexes; for each piece of searched data, the relevance of the query statement and the searched data is determined according to the data consistency index, the data integrity index and the data reliability index of the data source to which the data belongs, and the first N pieces of data with higher relevance are shown to the user. The problems of difficulty in integrating heterogeneous data, low processing and retrieval efficiency are solved.

This invention provides a lighting control method and system with self-detection upon power-on, relating to the field of lighting control technology. By selecting a reference node on the circuit board to calculate a signal integrity compensation coefficient, a current detection circuit is added to non-RGB lighting fixtures (12V powered headlights, low beams, high beams, etc.) to compare the actual operating current with the dynamic calibration current threshold to determine the status. For RGB LEDs (5V powered LEDs with communication functions, such as ambient lights), the red, green, and blue LEDs are controlled to light up sequentially, and feedback signals from the last LED are received. The signals are analyzed to determine the status of each LED, and finally, the two types of detection data are merged to form a comprehensive dataset, enabling fault identification, recording, and alarm functions. This invention improves the reliability of lighting control and the efficiency of fault diagnosis.

A system and method for generating synthetic data may include: receiving instructions to generate a synthetic data set having characteristics associated with a specified tumor methylation fraction; identifying, based on the received instructions, a first data set associated with an actual cancer sample and a second data set associated with an actual non-cancer sample; determining, based on the specified tumor methylation fraction, a first ratio of the first data set to be combined with a second ratio of the second data set; selecting, based on the determination, a first subset of the first data set corresponding to the first ratio and a second subset of the second data set corresponding to the second ratio; and using a processor to combine the first subset of the first data set and the second subset of the second data set to generate at least one synthetic data sample within the synthetic data set.

The present application relates to the technical field of biodiversity protection and ecological modeling, and particularly relates to a method and system for predicting hot spot distribution of large and medium-sized protected mammal diversity. According to the mammal diversity data obtained in the target area, spatio-temporal matching and standardization processing are performed to construct a comprehensive data set. By setting a variance inflation factor and through expert scoring method, core feature variables are selected from the comprehensive data set. Based on random forest and MaxEnt model as the basic prediction layer, the core feature variables are predicted, and the outputs of the random forest and MaxEnt model are respectively introduced into the meta learner model to obtain the initial distribution probability of each species. Habitat connectivity index and time series dynamic factor are constructed to generate a species diversity distribution matrix. A hot spot analysis algorithm is used to identify and output the hot spot area of each species. Through integration of multi-source data and dynamic correction mechanism, the accuracy and timeliness of the hot spot prediction are improved.

The present application relates to the field of artificial intelligence, and more particularly, to a method, an apparatus, a computer readable storage medium and a computer program product for generating synthetic data. The method comprises: obtaining original data and a condition set; training a generative adversarial network based on the original data and the condition set to generate initial synthetic data with conditional labels; dividing the initial synthetic data with conditional labels into a plurality of synthetic data sets; and adjusting the generative adversarial network using one or more of the plurality of synthetic data sets to generate final synthetic data.

This invention relates to the field of geophysical exploration technology and discloses a method for environmental risk assessment of geophysical mapping in groundwater-rich areas of subway construction zones. The method includes the following steps: S1, Data Acquisition and Fusion Step: Collecting ground-penetrating radar (GPR) data, surveying data, and engineering geological information from the subway construction area; spatially registering the GPR data and surveying data in a unified coordinate system to establish a comprehensive dataset containing the reflection characteristics of the subsurface medium and the spatial locations of surface and subsurface structures. By spatially registering and fusing GPR data, surveying data, and engineering geological information in a unified coordinate system, a comprehensive dataset containing the reflection characteristics of the subsurface medium and the precise spatial locations of surface and subsurface structures is established. This method overcomes the problem of disconnect between traditional geophysical interpretation and engineering spatial location, laying a reliable data foundation for subsequent accurate modeling and risk assessment.

A neural network system for retraining operational neural networks using a synthetic data set generated by a synthetic data generator neural network is provided. The synthetic data generator network comprises an input layer for receiving an input data set; an output layer for outputting the synthetic data set; and a loss function for receiving from each operational network a value of a medical metric. The operational networks each comprise an input layer for receiving the synthetic data set; and an output layer for outputting the value of the medical metric. The synthetic data generator network is trained for generating the synthetic data set based on the loss function comprising a difference of the values of the medical metric. Each operational network is retrained using the synthetic data set.

This invention discloses a data collection system and method for meteorological big data analysis, belonging to the field of meteorological data collection technology. It includes a data acquisition unit, a data processing unit, a data calibration unit, a fusion output unit, and a storage unit. By collecting historical meteorological data, real-time monitoring data, geospatial data, and equipment attribute data from multiple sources, and performing standard score conversion and interquartile range outlier removal, data quality and consistency are improved. The AdaCalib and NeuCalib algorithms are employed, combined with piecewise linear regression and neural networks for adaptive calibration, and dynamic adjustment and incremental learning are introduced to eliminate sensor drift and environmental interference, ensuring long-term monitoring accuracy. Time synchronization and WGS-84 coordinate mapping achieve spatiotemporal unification, integrating data into a comprehensive dataset containing meteorological parameters, timestamps, geographic coordinates, and equipment attributes, breaking down data silos, supporting in-depth analysis, and providing high-quality data support for precise meteorological services and decision-making.

The application relates to a highway slope photovoltaic glare evaluation method and device based on a UAV. The method comprises the following steps: controlling the UAV to fly along a flight path of a pre-planned observation point of a highway covering a photovoltaic module area and simulating a driver's visual angle, synchronously collecting multi-source data including images, reflection spectra, illumination intensity, a sun position and three-dimensional coordinate point clouds to form a comprehensive data set; automatically extracting key parameters of the photovoltaic module, such as an orientation, an inclination, surface cleanliness and reflection spectral characteristics, based on the images, the spectra and the point cloud data in the data set; further combining the sun position information and the illumination data of the observation point of the highway, and using a physical optical model to quantitatively calculate the glare intensity, the duration and the influence range of each observation point; finally, comparing the calculation results with preset threshold values, outputting intuitive glare grade evaluation results and spatial distribution information, and realizing accurate evaluation of the highway slope photovoltaic glare, so as to provide a scientific basis for ensuring driving safety and optimizing the photovoltaic system.

The present application belongs to the technical field of automatic driving perception, and particularly relates to a vehicle door state video editing method and system based on a diffusion model and physical priori. The method comprises: inputting a real driving video sequence and three-dimensional target detection and tracking label information thereof; screening a target vehicle based on the information, analyzing the visibility of the target vehicle in each frame and each perspective to determine an editing perspective; for each frame of each editing perspective, determining an editing region based on vehicle three-dimensional bounding box projection and constructing a text instruction; inputting an original video segment, frame-by-frame region information and the text instruction into a video diffusion model supporting mask editing to edit the vehicle door state in time sequence; based on a physical mapping relationship between a vehicle door size and an opening angle, automatically calculating an expansion distance, expanding a bounding box in a three-dimensional space and projecting to generate two-dimensional labels of each frame and each perspective; and finally constructing a synthetic data set. The present application solves the problems of vehicle door opening scene data scarcity and automatic generation of fine labeling.

This application relates to a method and system for collaborative delivery control of water rescue supplies based on unmanned aerial vehicles (UAVs). The method constructs a comprehensive dataset by real-time collection of the physiological state of the person in the water and parameters of the aquatic environment. It establishes a parameterized model by combining the performance of heterogeneous UAVs and the attributes of rescue supplies. A ternary matching model of supplies-UAV-target is used to solve a multi-objective optimization function to generate a Pareto optimal solution set. Based on spatiotemporal constraints, the optimal solution is analyzed to output a collaborative scheme including loading instructions, flight routes, and delivery timing. An incremental replanning mechanism triggered by environmental changes is introduced to achieve dynamic optimization of rescue resource allocation and closed-loop control of the delivery process in complex aquatic scenarios. This significantly improves rescue response speed and supply delivery accuracy, effectively reducing resource consumption and mission failure risks.

This application relates to a method, system, equipment, and medium for optimizing the operation mode of a power automation system. The method collects real-time operating data of the power system and health characteristic data of aging equipment to obtain a comprehensive dataset. Based on the comprehensive dataset, it calculates the health factors of each aging device and corrects the basic model parameters of system components, constructing a state-space model reflecting the aging characteristics of the equipment. On this basis, it constructs a multi-objective optimization problem with economic efficiency, reliability, and environmental protection as optimization objectives. Control variables are encoded to generate an initial population, and a non-dominated sorting genetic algorithm is used for iterative evolution to obtain a Pareto optimal solution set. Based on the Pareto optimal solution set, a multi-attribute decision-making method is used to select the comprehensive optimal operating mode, and after safety verification, it is executed. This achieves the optimal operating scheme that considers the impact of equipment aging through multi-objective optimization and decision-making screening, taking into account multiple dimensions of indicators, thereby improving the operational efficiency and scientific decision-making of the power automation system under aging scenarios.

The application discloses a kind of urinary calculus patient liquid intake intelligent management system, specifically related to medical health intelligent management technical field, including patient management data acquisition module, patient management data preprocessing module, liquid intake one-time management module, liquid intake secondary management module, intelligent management doctor-patient interactive module and intelligent management data storage module;The patient management data acquisition module is used to collect the multidimensional data of patient, including basic physiological data, real-time state data and diagnosis and treatment data, to obtain patient management first comprehensive data set;The application constructs the hierarchical control architecture that liquid intake one-time management and secondary management are combined, one-time management realizes basic dynamic precision control, secondary management is implemented differentiating reinforcement intervention to risk level, solve the problem that prior art management dimension is single, and the problem of insufficient pertinence, significantly improve management effectiveness.

The application relates to the technical field of intelligent agricultural equipment, and discloses a potato seeding machine and fertilizer adding vehicle path planning method and system, the method comprising the following steps: acquiring a comprehensive data set to predict a speed fluctuation trend; calculating a unit distance fertilizer consumption rate and an adjusted distance, and integrating to obtain a total consumption; judging a replenishment opportunity according to a difference value of a residual amount, and triggering path planning if the difference value is lower than a preset proportion; combining a fertilizer adding vehicle position and a seeder track to search for a potential intersection point with the minimum driving distance; evaluating downtime and fertilizer uniformity to determine an optimal intersection point; updating navigation instructions to generate a real-time route; monitoring a relative distance, activating a protocol to obtain a replenishment confirmation when entering a docking range; and feeding back an efficiency index to determine a next prediction starting point. Through path planning, the application realizes efficient and accurate collaborative operation of the seeder and the fertilizer adding vehicle in a complex field environment, and significantly improves operation efficiency and fertilizer utilization.

The application discloses an off-grid and grid-connected storage and charging switching management method and relates to the field of intelligent energy management. The method comprises the following steps: collecting data through installed sensors to obtain intelligent energy management comprehensive data, preprocessing the collected intelligent energy management comprehensive data to form an intelligent energy management comprehensive data set, transmitting the formed intelligent energy management comprehensive data set to a local edge computing node through wireless sensor network technology, obtaining a future energy demand change trend through time series analysis, generating a power adjustment strategy according to the future energy demand change trend, and adjusting the power by using a deep reinforcement learning model of an intelligent energy management panel. When a trigger condition is detected, a switching program is immediately started to convert off-grid to grid-connected. The application realizes comprehensive data coverage through data collection by installed sensors.

The application discloses a kind of based on streaming data's buried pipeline corrosion rate risk assessment method and system, the method includes: by collecting the multidimensional data of pipeline surrounding soil environment, pre-processing and isomeric data alignment, construct comprehensive dataset;Utilize streaming data processing framework to extract the dynamic change trend of environment and pipeline state, establish corrosion acceleration model;Depth analysis is carried out to key time node, determine corrosion diffusion evolution law;Optimize model parameters in combination with historical data, realize the adaptive prediction of corrosion speed under different environments;The application continuously inputs streaming data, corrosion risk grade is calculated in real time, and when exceeding threshold value, trigger early warning mechanism, generate targeted maintenance suggestion;By dynamically adjusting data acquisition strategy, fine monitoring to high-risk area is realized, to improve the accuracy and timeliness of buried pipeline corrosion prediction, provide strong support for pipeline maintenance decision.

The present application relates to a new energy project investment income and risk dynamic prediction method, belonging to the technical field of new energy project data analysis and intelligent decision-making. Among them, the method comprises: S1 obtains new energy project investment related multi-modal data and carries out standardization pretreatment, carries out adaptive fusion processing on the pretreated data, dynamically adjusts the fusion contribution degree to form a comprehensive data set; S2, based on the comprehensive data set, an integrated dynamic prediction model is constructed, a preset income and risk two-way coupling relationship is set, the income of green electricity and carbon assets is calculated, the risk quantification and stress test are carried out, and the multi-scale calculation results are output; S3, the deviation of model output and actual operation data is monitored in real time, the model is automatically triggered for updating, and the latest data is used for incremental adjustment; S4, the calculation results are integrated to evaluate the investment value and potential risk of the whole life cycle of the project, and the differentiated investment decision-making suggestions are generated. The linkage and integration of income and risk are realized, which can dynamically adapt to market and environmental changes.

The application discloses a boiler fly ash floating black treatment method, which comprises the following steps: quantitatively measuring the fly ash floating black of a boiler; arranging a sensor array, constructing a two-dimensional high-resolution monitoring grid covering the whole horizontal flue cross section, and acquiring two-dimensional oxygen content distribution data and two-dimensional temperature distribution data in real time; collecting actual operation data of the boiler, and jointly constructing a comprehensive data set by using the actual operation data, the two-dimensional oxygen content distribution data and the two-dimensional temperature distribution data; calculating key characteristic indexes including a wind-coal matching index, an oxygen content balance index and a temperature balance index based on the collected operation data; establishing a fly ash floating black prediction model to predict the fly ash floating black amount; and controlling the system to perform hierarchical closed-loop optimization control according to the output prediction floating black index and the calculated characteristic indexes. Through the closed-loop adjustment, the application realizes online prediction and accurate treatment of the fly ash floating black risk, and effectively reduces the fly ash floating black phenomenon.

The present application relates to the technical field of nodular cast iron, and particularly relates to a nodular cast iron full-process dynamic regulation method, which comprises collecting real-time data of temperature, pressure, weight and element concentration in the steps of batching, smelting and pouring to form a unified data set; smelting temperature and composition uniformity data are fused by using a data fusion algorithm to obtain a comprehensive data set; after meeting the integrity threshold, a parameter correlation graph is generated by association rule mining; key quality influencing factors are identified based on a decision tree algorithm; real-time feedback data and the correlation graph are compared to generate an adjustment instruction sequence; the instruction is executed and the production parameters are optimized through a feedback loop; product stability indicators are monitored, and when deviating from the preset range, the key influencing factors are updated; difference information is iteratively processed to obtain a final dynamic regulation model. The present application realizes full-process data collaboration and dynamic closed-loop regulation, improves product quality stability and production intelligence level, and is suitable for high-end nodular cast iron continuous production scenarios.

This invention discloses a deep learning-based method for co-optimization of cantilever widening and soil-rock coordination in mountainous highways. By acquiring geological survey and engineering monitoring data and fusing them from multiple sources, a comprehensive dataset containing soil-rock parameters, slope deformation, and load information is obtained. Subsequently, a generative adversarial network is used to deeply train this comprehensive dataset, accurately predicting the mechanical response of the soil-rock mass and the stress distribution of the cantilever structure. A co-optimization objective function is constructed with constraints of structural safety, ecological protection, and economic efficiency. A regression model is used to fit the quantitative mapping relationship between cantilever length, anchor bolt arrangement, and structural stiffness on soil-rock stability. Finally, a particle swarm optimization algorithm is used to iteratively search under multiple constraints to determine the optimal combination of the above parameters, which is then applied to the design of the cantilever structure. This invention significantly improves the design accuracy, safety, and overall benefits of cantilever widening projects under complex geological conditions.

The application discloses a method for accelerating a video diffusion model using a synthetic data set. The method comprises: generating a synthetic data set using a pre-trained video diffusion model, the synthetic data set containing a synthetic video, a denoising trajectory in a latent space, and a corresponding text prompt; using the pre-trained video diffusion model as a teacher model and constructing a corresponding student model, the student model and the teacher model sharing the same structure; based on the synthetic data set, performing knowledge distillation training on the student model, in the knowledge distillation training process, the student model learns the denoising process of the teacher model and aligns the data distribution generated by the teacher model until a set loss function standard is met; and using the knowledge distillation trained student model as a video generation model and applying it to a video analysis task. By using the application, a video with higher quality and higher resolution can be generated.