118 results about "Frequency data" patented technology

The invention belongs to the technical field of machine room power distribution, and provides a machine room power distribution equipment remote monitoring method and system, and the method comprises the steps: collecting operation data in real time through deploying edge equipment, fusing basic operation data, transient feature data and historical abnormal data, and constructing a state scoring model; utilizing a neural network training weight coefficient to realize equipment health degree dynamic evaluation; the sampling frequency is adaptively adjusted based on a scoring result, abnormal data are screened by combining multiple characteristic thresholds such as current sudden change intensity, energy sudden change quantity and temperature gradient, and only high-frequency data fragments in an abnormal time period are reserved; through an edge-cloud collaborative storage mechanism, abnormal data are locally cached after RAID redundancy check, normal data are uploaded to the cloud according to storage duration and access frequency in a grading manner, full-life-cycle traceability of the data is realized through tagging processing, anomaly detection sensitivity is improved, and storage efficiency and fault diagnosis reliability are both considered.

The invention relates to the field of sensor data processing, and discloses an MEMS drift correction method based on data dynamic sampling and multi-source fusion, and the method comprises the following steps: S1, collecting MEMS three-axis acceleration original data at a first sampling frequency, and synchronously collecting environment temperature, humidity and air pressure data at a second sampling frequency, the first sampling frequency is higher than the second sampling frequency, the basic drift correction model is established by using the low-frequency data to capture the macroscopic and slow-varying relationship between the environmental factors and the drift, then the residual refinement model is established for the high-frequency data, and the high-frequency residual is specially processed; high-frequency drift components related to the environmental change rate are captured, through the two-stage fusion strategy, the accuracy of the model in the macroscopic trend is guaranteed, high-frequency data are fully utilized to improve the instantaneous correction precision, information loss caused by simple downsampling in a traditional method is avoided, and false signals introduced by rough interpolation are also avoided.

The invention provides an oil well load indicator diagram prediction method and system, and the method comprises the steps: synchronously collecting an initial electrical parameter data set, a load data set, an accelerometer data set and a barometer data set of an oil well, the initial electrical parameter data set comprising an active power data unit, and then obtaining an active power data segment and a load data segment; obtaining a time lag based on the two, and further obtaining a final electrical parameter data segment; acquiring a reference displacement sequence through the accelerometer data set and the barometer data set; obtaining a dynamic stroke length based on the reference displacement sequence and the load data segment, and further obtaining a final displacement sequence; and constructing a prediction neural network, and obtaining an oil well load indicator diagram based on the prediction neural network and the real-time electrical parameter data segment. According to the technical scheme, the indicator diagram is obtained through continuous and high-frequency real-time electric parameter data segments, motion period information loss caused by low-frequency data collection is avoided, accumulative errors are avoided, and the accuracy of the indicator diagram is ensured.

The invention provides a cooling tower system optimal operation parameter self-adaptive regulation and control method, which comprises the steps of collecting multi-source high-frequency data in a system operation process, and obtaining a standardized variable sequence through normalization, abnormal value processing and time sequence consistency verification. A sliding window mechanism and a lightweight causal discovery algorithm are combined, a causal atlas is dynamically constructed and fused, parameters are corrected in real time through an incremental Bayesian network, a dominant causal path is extracted, lightweight CMA-ES rolling optimization is executed in an equivalent control subspace after dimension reduction, optimal control parameters are solved, system closed-loop adjustment is driven, and the optimal control parameters are obtained. Meanwhile, a map confidence degree evaluation and depth reconstruction mechanism is set, dynamic self-adaption of a causal structure and parameters is ensured, the response capacity of a cooling tower system to working condition changes can be improved, and energy efficiency and operation stability are optimized.

The invention relates to the technical field of power system frequency stability control, in particular to a power system inertia online evaluation method and device, equipment and a medium, and the method comprises the steps: obtaining frequency deviation data and active power data of each node, carrying out the low-pass filtering of the frequency data, and calculating the amount of active unbalance; an inertia response model is constructed based on the rotor motion equation of the synchronous generator, and inertia and damping coefficients are estimated by using a least square method; constructing a transfer function and judging the stability of the transfer function, if the transfer function is stable, inputting an active unbalance amount to obtain a frequency deviation predicted value, and calculating a Pearson correlation coefficient with an actual frequency; different time periods are selected through a sliding window to repeat the process, finally, the inertia corresponding to the maximum Pearson's correlation coefficient is selected as a system inertia evaluation result, the inertia level of the system can be accurately evaluated on line through the method, and therefore data support is provided for frequency control, and the frequency stability and safety of the system are guaranteed.

The invention relates to the technical field of information retrieval, in particular to an intelligent screening method based on multi-source data, which comprises the following steps: acquiring high-frequency time sequence data of a first data source, and dividing the high-frequency time sequence data into a state change interval and a state stable interval; extracting dynamic features from the state change interval, extracting statistical features from the state stable interval, and generating a pattern feature cluster sequence; acquiring low-frequency data of a second data source, acquiring a timestamp as an anchor point, acquiring a precursor node and a subsequent node of a mode feature cluster in combination with the mode feature cluster sequence, and generating an anchor point association map; executing bidirectional prediction verification based on the anchor point association graph, calculating a bidirectional consistency score, and generating an alignment confidence label; and monitoring incremental updating of high-frequency time sequence data, performing reverse prediction verification again, updating and aligning confidence labels, screening anchor point association records with high confidence, performing fusion processing on low-frequency data features and mode feature clusters, and outputting a screening result.

The invention relates to the technical field of data processing, in particular to a user and entity behavior modeling anomaly detection system, which comprises a timestamp generation module for receiving user power consumption data, equipment monitoring data and power grid monitoring data to generate data vectors; the dynamic association module interpolates the low-frequency data in the same time window to generate a virtual sampling point, and outputs a multi-dimensional association tensor; the behavior model generation module generates a joint behavior model and a confidence probability; the detection decision-making module adopts four parallel optimization channels to output three-dimensional decision-making indexes including abnormal probabilities; the simulation optimization module drives multi-algorithm simulation in the digital twin environment and writes back optimal parameters; and the feedback correction module generates a time reversal data flow reconstruction behavior model based on the simulation error, and outputs an equipment binding early warning signal to a power grid safety system. According to the method, the problem of dynamic association missing caused by multi-source data time sequence asynchronization is solved, and efficient collaborative diagnosis of composite anomalies such as electricity stealing is realized.

The invention discloses an electric power real-time dynamic carbon factor metering and monitoring system, which belongs to the technical field of electric power systems, and comprises a data acquisition module used for acquiring multi-source data of a power generation side, a power grid side and a user side of an electric power system in real time, verifying the multi-source data, generating high-frequency data streams, and transmitting the high-frequency data streams to a data processing module; the high-frequency data stream comprises renewable energy output fluctuation data, load change data and power grid topological structure data; the dynamic calculation module is used for generating time-sharing and partitioned dynamic carbon factors based on the high-frequency data stream; the monitoring feedback module is used for generating a carbon flow path track based on the dynamic carbon factors and generating a calibration instruction set in combination with deviation records in the historical dynamic carbon factors; and the optimization output module is used for recalculating the dynamic carbon factor based on the calibration instruction set and the real-time multi-source data, and outputting the optimized dynamic carbon factor of the time-sharing partition to an external carbon accounting system. According to the method, the problem that the traditional annual average factor cannot reflect the space-time dynamism is solved by calculating the carbon factor of the time-sharing partition.

The invention discloses a health detection instrument data processing system and method based on the Internet of Things, and relates to the technical field of data analys.The method comprises the following steps that sampling intervals are determined, a CGM sensor is controlled to collect data according to the corresponding sampling intervals, and a frequency mixing blood glucose data set is formed; constructing an input data set, and introducing a sampling interval weight to construct a basic trend model; based on the input data set, introducing a correction item to construct a scene correction model, and outputting a final prediction value at the current next moment; setting a future prediction time interval and a prediction moment number N, generating N prediction values based on the basic trend model and the scene correction model, and integrating the N prediction values to form a prediction sequence; setting a threshold value to analyze whether the prediction sequence is abnormal or not and the abnormal type; when an anomaly is detected, a feedback mechanism is triggered. The method can effectively improve the situation that in the prior art, dynamic blood glucose monitoring sampling frequency lacks scene adaptability and is insufficient in frequency mixing data processing capacity.

The invention discloses a frequency data analysis method and device based on self-supervised learning, and a storage medium. The method comprises the following steps: S1, constructing a standard spectrum tensor; s2, executing frequency change feature extraction operation to generate a frequency change feature map; s3, determining a frequency section of which the hopping intensity is greater than a preset hopping threshold value, constructing a hopping section mask, and generating a structure covering spectrum tensor; s4, inputting to an improved ConvTransform model, and generating a frequency spectrum reconstruction tensor and a residual feature tensor; s5, constructing a frequency spectrum reconstruction loss function and a residual error fitting loss function, executing a self-supervised training process, and constructing a frequency guidance feedback graph; and S6, applying to frequency data analysis, and generating a frequency spectrum reconstruction result, a residual error response result and a hopping section positioning result. According to the invention, the abnormal change identification capability and the frequency spectrum reconstruction precision in the frequency data analysis process are improved.

This invention discloses a high-frequency macroeconomic monitoring method and system based on an intelligent model. The method constructs a high-dimensional time-varying mixed-frequency dynamic factor model to establish a dynamic factor structure of time-varying parameters for daily, low-frequency flow, and stock indicators. It also constructs a real-time GDP tracking model, decomposing quarterly GDP into daily GDP and using it as a common factor, with other indicators providing updated estimates through time-varying regression. Joint parameter estimation is performed using a quasi-maximum likelihood method and Kalman filtering. In application, the model parameters are dynamically updated using real-time released mixed-frequency data, ultimately extracting the daily economic condition index and the daily GDP growth rate sequence, achieving multi-frequency real-time macroeconomic monitoring and early warning. This invention improves the timeliness, adaptability, and accuracy of macroeconomic monitoring.

The invention belongs to the technical field of arc detection, and relates to an aviation alternating current series arc fault detection method based on a variable coefficient, which comprises the following steps: collecting a current signal, dividing the current signal into a continuous time window, and forming a current frequency data set; performing frequency component analysis on the current frequency data set to obtain a frequency distribution vector, and calculating a standard deviation and a mean value of the frequency distribution vector to obtain a variable coefficient; forming a variable coefficient sample set based on historical current data collected under different loads, and performing dual Weibull distribution fitting; selecting a numerical value of a variable coefficient corresponding to a preset quantile level as an adaptive threshold value; and if the variable coefficient of the current time window is greater than the adaptive threshold, determining that an alternating current series arc fault exists, and outputting alarm information. According to the technical scheme of the invention, the current behavior is represented in a form capable of reflecting the characteristics of the frequency structure, and the stable recognition under the multi-load and disturbance working conditions is realized through the variable coefficient and the adaptive threshold.

The present disclosure provides a business test data extraction method applied to U shield testing, which can be applied to the fields of big data technology and artificial intelligence technology. The business test data extraction method applied to U shield testing comprises: generating a high-frequency data set and a low-frequency data set corresponding to a plurality of test U shields according to the historical use frequency of each test U shield in the plurality of test U shields corresponding to a business category; updating the high-frequency data extraction rule corresponding to the business category at the current time through the high-frequency data features corresponding to the high-frequency data set and the low-frequency data features corresponding to the low-frequency data set; and extracting the corresponding business test data based on the high-frequency data extraction rule to realize the U shield testing. The present disclosure also provides a business test data extraction device, equipment, storage medium and program product applied to U shield testing.

An object of a system according to an embodiment is to propose an optimal delivery timing based on a use frequency of a user.SOLUTION: A system includes a collection unit, an analysis unit, and a proposal unit. The collection unit collects use frequency data of a user. The analysis unit analyzes the data collected by the collection unit and calculates a delivery timing. The proposing section proposes a delivery timing to the user based on the analysis result obtained by the analyzing section.SELECTED DRAWING: Figure 1

The invention discloses a real-time data processing method and system based on frequency layering, and belongs to the technical field of computers, and the method comprises the steps: obtaining a layering configuration rule and a preset processing period of a data table according to attribute frequency; acquiring a corresponding change attribute according to the change of the real-time data, and acquiring a change plane and change time to which the change attribute belongs according to a hierarchical configuration rule; if the change attribute belongs to a frequent change plane in the hierarchical configuration rule, entering a frequent change increment queue and processing the change attribute and the change time according to the change time and a first period in a preset processing period, and if the change attribute belongs to a non-frequent change plane in the hierarchical configuration rule, entering a non-frequent change increment queue and processing the change attribute and the change time according to the non-frequent change plane in the hierarchical configuration rule; and processing the change attribute and the change time according to the change time and a second period in the preset processing period. The technical problem of low data processing efficiency caused by the fact that high-frequency data items and low-frequency data are not layered and distinguished in the prior art is solved.

This application relates to a wheel detection method, apparatus, computer equipment, and storage medium. The method includes: acquiring noise time-domain data of the target wheel at a preset acquisition frequency using a noise acquisition component corresponding to the target wheel; determining the dominant noise frequency data of the target noise time-domain data in a preset frequency band based on the target noise time-domain data, and determining the target wear order of the target wheel based on the dominant noise frequency data; determining the target sound pressure level density threshold corresponding to the target wear order based on the target wear order and a preset first correspondence between the wear order and a sound pressure level density threshold; and generating detection information indicating wheel abnormality if the dominant frequency sound pressure level density in the noise frequency data is greater than or equal to the target sound pressure level density threshold. This solution allows for real-time detection of the target wheel.

The application relates to an intelligent data classification storage method and system based on deep learning, which comprises the following steps: obtaining an original data set, and obtaining a multidimensional data set after preprocessing; obtaining a frequency statistical set of data, and fitting the frequency statistical set by using a least square method to obtain an access frequency prediction value of each type of data; generating high-frequency or low-frequency data identification for corresponding data according to the comparison between the access frequency prediction value and a preset frequency threshold; and constructing a high-frequency data list and a low-frequency data list accordingly; obtaining a storage cost weight and an access delay index of a storage node according to a storage location index; calculating a storage efficiency score of each storage node; calculating an allocation proportion of each storage node, adjusting the allocation proportion through the storage efficiency score, and generating a multi-level storage allocation scheme; and after generating corresponding migration instructions according to the multi-level storage allocation scheme, migrating data to realize hierarchical classification storage of the data. The application can improve the utilization rate of system storage resources.

To provide an EMC countermeasure presentation system capable of efficiently extracting and presenting an emission countermeasure case in another field and another model with high similarity (countermeasure effectiveness) in consideration of a device configuration of an electronic apparatus.SOLUTION: A similar configuration extraction unit 3 configured to extract, based on device configuration data 11 at the time of noise measurement, device configuration data at a short distance by configuration learning data 9 obtained by learning a graph model representing a device configuration when the noise countermeasure and reduction amount frequency data is acquired; and a countermeasure estimation unit 4 configured to estimate, from data obtained from the reduction countermeasure extraction unit and the similar configuration extraction unit, a recommended countermeasure content having a high degree of similarity of a device connection configuration and expected to have an excessive noise reduction effect.SELECTED DRAWING: Figure 1

The invention relates to the technical field of economic measurement modeling, in particular to an economic prediction method and system based on weekly, monthly and quarterly big data, electronic equipment and a medium. The method comprises the following steps: acquiring multi-frequency multi-source economic time sequence data; performing frequency mixing alignment preprocessing on the multi-source economic time series data according to a preset frequency mixing data processing rule to generate a frequency mixing data set; constructing a connection model through a predefined modular architecture and a connection algorithm based on the frequency mixing data set; training the connection model through a parameter estimation algorithm; and inputting multi-source economic time sequence data obtained in real time into the trained connection model, executing high-frequency prediction, and outputting a real-time prediction result. By means of the mode, the technical problem that when an existing mixing dynamic factor model deals with multi-frequency mixing data, prediction precision is insufficient is solved, and the capacity of conducting high-frequency, accurate and real-time monitoring on the macroeconomic trend is improved.

The application belongs to the technical field of arc detection, and relates to an aviation alternating current series arc fault detection method based on a variation coefficient, comprising the following steps: collecting a current signal, dividing the current signal into continuous time windows, and forming a current frequency data set; performing frequency component analysis on the current frequency data set to obtain a frequency distribution vector, calculating a standard deviation and a mean value, and obtaining a variation coefficient; forming a variation coefficient sample set based on historical current data collected under different loads, and performing double Weibull distribution fitting; selecting a numerical value of the variation coefficient corresponding to a preset quantile level as an adaptive threshold value; if the variation coefficient of a current time window is greater than the adaptive threshold value, it is determined that an alternating current series arc fault exists, and alarm information is output. The technical scheme of the application characterizes the current behavior in a form capable of reflecting frequency structure characteristics, and realizes stable identification under multiple loads and disturbance working conditions through the variation coefficient and the adaptive threshold value.

The application discloses an intelligent monitoring system for Eriocheir sinensis breeding environment, and particularly relates to the technical field of intelligent monitoring of aquaculture, which collects data through multiple climbing cage monitoring modules, calculates average weight according to the quantity and weight of image recognition and corrects the climbing cage frequency, carries out spatial analysis based on the corrected frequency to obtain a spatial aggregation global index and a local aggregation index, fuses to generate a spatial aggregation degree characteristic value, compares and calculates a deviation degree index by a deviation calculation module, and selectively starts the regulation and control equipment of different spatial regions according to the deviation degree index size by a regulation and control execution module; the climbing cage frequency is corrected through image and weight data, the single semantic problem of frequency data is solved, the spatial aggregation degree characteristic value is compared with the historical benchmark, the dynamic correlation between the environment and the behavior is realized, the deviation degree index is classified and partitioned for regulation and control, the execution lag problem is solved, and the precise intervention and energy consumption optimization are realized.

The invention discloses a real-time all-weather risk control system based on high-frequency data and news public opinions, and the system comprises a quantitative intelligent transaction module, a risk income calculation module, a multi-asset risk control module, and an intelligent evaluation module. The quantitative intelligent transaction module comprises a strategy combination module, a strategy matching fund module, an automatic transaction module and a strategy compliance real-time review sub-module, the sub-module obtains financial supervision compliance requirements in real time, carries out compliance verification on financial product combination and fund putting schemes, blocks execution of uncompliance rules, and feeds back the uncompliance rules; the risk income calculation module comprises a core data collection module and a data traceability evidence storage sub-module, the sub-module marks a source identifier, collection time and personnel information for core data to form non-tampering traceability chain associated storage, and the multi-asset risk control module comprises a risk control module and a risk slow release scheme automatic generation sub-module. The sub-module receives the risk signal and analyzes the risk type and range.

The invention discloses a PLC high-frequency data tracing system and a fault diagnosis method, and relates to the technical field of data tracing and fault diagnosis. PLC related high-frequency data tracing and fault diagnosis interfaces are configured, PLC working information is obtained, PLC internal variable data designated by a user are collected based on the PLC working information, and the PLC internal variable data are stored in a database. The method comprises the following steps of: judging whether a fault occurs in the operation of the PLC, locking a data buffer area when the fault occurs, forming a data snapshot related to the fault, uploading the data snapshot to a diagnosis and analysis module, and carrying out visual analysis on the uploaded snapshot data. The system can be safely deployed on a key production line which is running for a long time, does not need to worry that the system becomes a new interference source, can quickly and accurately position and solve accidental faults, effectively reduces the average fault repair time of equipment, and directly improves the comprehensive efficiency of the equipment.

The application provides an electronic transformer high-frequency data analysis and storage method and system. The method comprises the following steps: calculating coarse amplitude information, coarse phase information, fine amplitude information and fine phase information of measurement data according to message data output by an electronic transformer; realizing electronic transformer measurement data type identification by performing feature analysis on the coarse amplitude information and the coarse phase information in a short time domain, and judging whether the electronic transformer measurement data is steady-state data or transient-state data; realizing differentiated data storage according to the electronic transformer measurement data type identification, taking a long-time coarse storage method for the steady-state data, and taking a short-time fine storage method for the transient-state data. The application solves the technical problem that the analysis and storage method of the electronic transformer measurement data cannot adapt to various operating states of the electronic transformer, resulting in insufficient storage dimensions of the electronic transformer measurement data and the inability to support online evaluation of the operating state of the electronic transformer.

The present application relates to a kind of seismic waveform classification calculation methods based on multi-frequency data body, belong to the high-quality reservoir prediction technical field of energy exploration field such as oil and gas field exploration, shale gas exploration, geothermal exploration, the method is first based on the 3D seismic data of target interval in study area, using time-frequency analysis method calculates 3 common frequency data bodies such as low frequency, medium frequency, high frequency, then sequentially carries out waveform clustering analysis method based on neural network training for 3 common frequency data bodies and carries out the binary analysis of seismic waveform, the study area is divided into 8 sub-regions, then using original data body carries out waveform overall correlation analysis and merging of 8 sub-regions, finally using different color scale represents the value of sub-region, drawing, using image representation waveform classification result.

The invention provides a parameter extraction and alignment processing method and system for DLCP test data, and the method comprises the steps: carrying out the DLCP test at a plurality of different test frequencies, and obtaining the capacitance bias curve data corresponding to each test frequency; carrying out validity judgment on the capacitance bias curve data, respectively carrying out function fitting on the curve data which are judged to be valid, and extracting capacitance parameters corresponding to the test frequencies from a fitting result; based on each capacitance parameter, calculating a spatial position parameter corresponding to each test frequency through a first model; performing alignment processing by taking a spatial position parameter corresponding to a preset high frequency as a reference, and outputting a multi-frequency data set under a unified depletion region width coordinate system; and performing multi-frequency capacitance data difference calculation to obtain trap density distribution, and outputting trap density space distribution data. The accuracy and the stability of a test result are improved by carrying out space alignment processing based on a unified physical model on the capacitance bias voltage data under different test frequencies.

The embodiment of the invention discloses a power grid operation state early warning method and device based on a sudden change prediction model, and the method comprises the steps: obtaining the operation data of a power grid, and the operation data comprise voltage data, frequency data, load data and meteorological parameters; constructing a power grid state nonlinear potential function model based on the operation data, wherein the potential function model is used for describing the stability of a power grid; analyzing the potential function model by using a mutation theory, and extracting a critical condition between a control variable and a state variable; identifying a critical area of the power grid operation parameters according to the critical condition, and establishing a corresponding relation between parameter change and power grid stability; and when the operation state of the power grid is close to the critical region and the power grid is evaluated to be in an unstable critical state according to the corresponding relation, generating early warning information. According to the invention, the safety and stability of power grid operation are improved.

This invention discloses a method for obtaining the quality factor based on stable trend frequency, comprising the following steps: S1 acquiring first arrival records, which include time-domain data and first arrival information; S2 lengthening the time-domain data and converting the lengthened time domain into a frequency domain to obtain the spectrum of the first arrival records; S3 extracting first frequency data from the spectrum and obtaining second frequency data based on the variation characteristics of the first frequency data; S4 calculating the quality factor Q value based on the second frequency data and the first arrival information. This invention also discloses a device, equipment, and medium for obtaining the quality factor based on stable trend frequency. This invention obtains a highly accurate Q value by lengthening the first arrival data, employing a stable trend frequency method, and optimizing the Q-value calculation algorithm. This method is not only applicable to obtaining Q values ​​from dual-well micrologging data but can also be extended to obtaining Q values ​​from seismic single-shot records.

The invention discloses a management method for computer abnormal behavior recognition, and relates to the technical field of computers, and the method comprises the steps: obtaining login information of a current login behavior of a user; matching a database according to account information in the login information to obtain a historical login data set; judging a login mode, and if the login mode is account retrieving type login, determining frequency data of the corresponding login mode in a historical login record; determining time difference data between the last login behavior and the current login behavior, and comparing the time difference data with the frequency data to obtain a first abnormal value; performing resource data monitoring on the logged-in account, determining the reasonability during resource scheduling, and obtaining a second abnormal value; and determining the security of the current login behavior according to the first abnormal value and the second abnormal value, and marking the unsafe current login behavior as an abnormal login behavior. The method has the effect of improving the recognition accuracy of the abnormal behaviors of the computer.

The application provides an oil well load work graph prediction method and system, the method comprising: synchronously collecting initial electric parameter data set, load data set, accelerometer data set and barometer data set of the oil well, the initial electric parameter data set comprising active power data unit, and then obtaining active power data segment and load data segment; obtaining time lag amount based on the two, and then obtaining final electric parameter data segment; obtaining reference displacement sequence through the accelerometer data set and the barometer data set; obtaining dynamic stroke length based on the reference displacement sequence and the load data segment, and then obtaining final displacement sequence; constructing a prediction neural network, and obtaining the oil well load work graph based on the prediction neural network and real-time electric parameter data segment. The work graph is obtained by using continuous and high-frequency real-time electric parameter data segment, so that the motion cycle information loss caused by low-frequency data collection is avoided, and then the cumulative error is avoided, and the accuracy of the work graph is ensured.