RTU (Remote Terminal Unit) data processing method and device for wide-area measurement time delay

Abstract

The invention relates to the technical field of power systems, in particular to an RTU data processing method and device for wide-area measurement delay, and the method comprises the following steps: 1, constructing a data storage structure; step 2, setting a storage time reference; step 3, dividing a storage area driven by two factors; step 4, carrying out time-sequence number synchronous conversion on the RTU and the scheduling master station; step 5, performing differential data updating; step 6, collecting and predicting measurement data; and 7, initializing and circularly updating. According to the method, a refined data storage structure is constructed, storage areas are divided according to equipment types and data features, storage resource configuration is optimized, and the storage cost is reduced while the access speed of important data is guaranteed.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of power systems, in particular to an RTU data processing method and device for wide-area measurement time delay. BACKGROUND

[0002] In the wide-area measurement of power systems, the remote terminal unit (RTU) as a key data acquisition device, the data collected by it needs to be transmitted to the dispatching master station in real time to support system monitoring and decision-making. However, due to factors such as communication link delay and device response difference, the RTU data often has transmission delay problems, which causes the data received by the dispatching master station to be out of synchronization with the actual measurement time. At the same time, the types of power system equipment are various, and the scale of measurement data is huge. The traditional data storage structure is difficult to balance storage efficiency and access speed, and the data update rule lacks pertinence, which is easy to cause data distortion or update lag, and abnormal data will also interfere with subsequent analysis and prediction, seriously affecting the accuracy and reliability of power system dispatching.

[0003] In view of the above problems, we propose an RTU data processing method and device for wide-area measurement time delay. SUMMARY

[0004] The purpose of the present application is to provide an RTU data processing method and device for wide-area measurement time delay to solve the problems raised in the background art.

[0005] To achieve the above purpose, the present application provides the following technical scheme:

[0006] An RTU data processing method for wide-area measurement time delay, characterized in that it comprises the following steps:

[0007] Step 1, data storage structure construction: constructing a data storage structure suitable for multiple types of power system equipment, establishing an independent storage table corresponding to the equipment type, and using a two-dimensional array to store the measurement data, wherein the rows of the two-dimensional array correspond to the equipment identifier and the columns correspond to the storage point serial number;

[0008] Step 2, storage time reference setting: setting a storage time reference, generating an absolute time corresponding to the storage point serial number by calculation, and constructing a discrete time sequence to determine the time anchor point of data storage;

[0009] Step 3, storage area division driven by double factors: dividing the storage area, based on the double factor calculation results of the equipment importance coefficient and the data activity index, dividing the storage resources into core storage area, ordinary storage area and edge storage area, the equipment importance coefficient is determined in combination with the equipment power level and the fault influence degree, and the data activity index is determined in combination with the data variation and the update interval;

[0010] Step 4: Time-Sequence Number Synchronization Conversion between RTU and Scheduling Master Station: Perform time-storage point sequence number conversion, standardize the RTU acquisition time to a second-level timestamp, and calculate the integer or non-integer storage point sequence number through the mapping relationship between the timestamp and the absolute time to achieve time synchronization between the RTU and the scheduling master station.

[0011] Step 5, Differentiated Data Update Execution: Perform differentiated data updates. Based on the RTU data refresh status and storage point sequence number type, trigger the corresponding update operation. The update operation includes weighted calculation update, keeping the original data unchanged, synchronous overlay update, and classification-triggered prediction update.

[0012] Step 6, Measurement Data Collection and Prediction: Collect and optimize measurement data, use data prediction algorithms to fill in missing data, identify abnormal data through residual calculation and perform replacement and repair, forming a complete and reliable measurement dataset;

[0013] Step 7, Initialization and Cyclic Update: Complete system initialization and cyclic update. When the system starts, it will execute the corresponding initialization correction strategy based on historical data. Subsequently, it will continuously execute the operations from Step 4 to Step 6 to achieve dynamic data maintenance.

[0014] Preferably, in step 1, the power system equipment is divided into 10 categories: circuit breakers, disconnectors, generators, loads, busbars, capacitors, AC line segments, transformer windings, converters, and DC line segments. Each category of equipment table contains a unique equipment identifier (ID), basic equipment parameters, and a measurement data storage index.

[0015] The two-dimensional array structure for storing measurement data is defined as follows: The row dimension corresponds to the unique identifier of the device, and the column dimension corresponds to the sequential number of the data storage point of the scheduling master station. Indicates the number of time divisions. To maximize the storage time, It is a fixed time interval.

[0016] Preferably, in step 2, the absolute time corresponding to the storage point number is determined by the formula... It is confirmed that, among them, At the initial time point, It is a time interval. The sequence number of the time step ( ), Then it represents the first The time values ​​corresponding to each time step are accumulated by multiples of the initial time and the time interval to obtain the values ​​of each time point in the discrete time series.

[0017] Preferably, in step 3, the formula for calculating the equipment importance coefficient is as follows: ,in a device power level weight, a device failure influence coefficient, and a weight coefficient of and respectively, and ;

[0018] The calculation formula of the data activity index is , wherein is the calculation result of the th measurement point, is the data change amount in a unit time, is the data update interval, is a normalization coefficient.

[0019] Preferably, in step 3, the storage area division is specifically:

[0020] Core area: when and , a high-speed storage medium is used;

[0021] Normal area: when or ;

[0022] Edge area: when and , a large-capacity storage medium is used.

[0023] Preferably, in step 4, the timestamp standardization is:

[0024] The conversion formula is , wherein represents the minute value, represents the second value;

[0025] Integer sequence number calculation:

[0026] The formula is , wherein is the time value at a certain moment, is the time period, is the modulo operator, is the time interval, is the rounding function;

[0027] Non-integer sequence number calculation:

[0028] The formula is , wherein is the floating-point value.

[0029] Preferably, in step 6, the telemetry interpolation update is:

[0030] Formula is Wherein And Respectively, the absolute time of the storage point And ; The historical measurement value is The current RTU refresh value is

[0031] The sliding window prediction model adopts an autoregressive moving average model Wherein The autoregressive coefficient is The moving average coefficient is , The model order of autoregression and moving average is The constant term is The white noise sequence is

[0032] Abnormal data processing:

[0033] Residual error Threshold , The standard deviation of the historical residual error is The confidence coefficient is , And Respectively, the adjacent normal data points are

[0034] Preferably, in step 7, the differential initialization strategy is:

[0035] When there is no historical data, execute Wherein The first RTU sampling value is The initialization correction coefficient is

[0036] When there is historical data, a time decay strategy is adopted , The time decay factor is

[0037] A wide-area measurement delay RTU data processing device, comprising:

[0038] The first processing module: when the RTU measurement data is not refreshed and the storage point serial number is an integer, the previous storage point measurement data in the scheduling master station real-time library is updated by weighted calculation to update the current storage point data, with an adaptive adjustment mechanism;

[0039] The second processing module: when the RTU measurement data is not refreshed and the storage point serial number is a non-integer, the real-time library data remains unchanged, the time delay deviation is monitored and the warning and calibration request is triggered;

[0040] The third processing module: when the RTU measurement data has been refreshed and the storage point serial number is an integer, the refreshed data is synchronously updated to the corresponding storage point of the real-time library, and parallel processing technology is adopted;

[0041] The fourth processing module: when the RTU measurement data has been refreshed and the storage point serial number is a non-integer, the remote signaling quantity is directly updated to the corresponding storage point, the remote measurement is updated to trigger the prediction process, and an intelligent decision system is integrated;

[0042] The data storage structure construction module: a multi-section time wide-area measurement storage structure is constructed, a device table is established according to the device type, the measurement data is stored by adopting a two-dimensional data structure, and the storage partition is adjusted by using an intelligent storage strategy;

[0043] The measurement data collection module: the measurement value is updated by using an interpolation algorithm, the first N data in the sliding window are input into an autoregressive moving average prediction model to calculate a predicted value, and a data quality monitoring and repairing mechanism is integrated.

[0044] Compared with the prior art, the beneficial effects of the present application are:

[0045] The present application constructs a refined data storage structure, divides the storage area according to the device type and data characteristics, optimizes the storage resource configuration, guarantees the important data access speed, and reduces the storage cost.

[0046] The time-serial number conversion mechanism and the differential data updating rule effectively solve the RTU data time delay problem, and improve the time synchronization and accuracy of the data.

[0047] The sliding window prediction model and the abnormal data processing mechanism are integrated, the data integrity and reliability are enhanced, and strong support is provided for power system dispatching decision.

[0048] The modules of the device work cooperatively, have the functions of adaptive adjustment, parallel processing, intelligent decision and the like, can efficiently process large-scale measurement data, and meet the needs of the wide-area measurement scene. BRIEF DESCRIPTION OF DRAWINGS

[0049] Figure 1 The figure is a schematic diagram of the system module of the present application. DETAILED DESCRIPTION

[0050] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the protection scope of the present application.

[0051] A RTU data processing method for wide-area measurement time delay, comprising the following steps:

[0052] Step 1, data storage structure construction;

[0053] Divide the device types and establish a device table:

[0054] The power system devices are divided into 10 types of circuit breakers, switches, generators, loads, buses, capacitors, AC line sections, transformer windings, converters and DC line sections, and an independent device table is established for each type of device. The device table fields include device unique identifier ID, device basic parameters and measurement data storage index.

[0055] Construct a two-dimensional measurement data storage structure:

[0056] A two-dimensional array structure is used to store measurement data in each device table. In the device data table, the storage structure is defined as , wherein the row dimension corresponds to the device unique identifier , and the column dimension corresponds to the sequential number of the dispatch master station data storage point , , wherein represents the time divided by the time interval to obtain the number; wherein is the calculated numerical result, represents the maximum storage duration, is the fixed time interval, which is commonly used in time series analysis, sampling calculation and other scenarios. The number of contained in the time is calculated by this formula.

[0057] Step 2, storage time reference setting;

[0058] The absolute time corresponding to the storage point number is determined by the formula , wherein is the starting reference time (time accuracy to seconds), and the value is limited to the range of 00:00:00-23:59:59, and the time sequence relationship of adjacent storage points is satisfied .

[0059] represents the calculation relationship of discrete time series, wherein is the initial time point, is the time interval, is the serial number of the time step ( ), then represents the The time value corresponding to each time step is obtained by accumulating the initial time and the multiple of the time interval, to get the numerical value of each time point in the discrete time sequence.

[0060] Step 3, storage area division driven by double factors:

[0061] , which means By and Two variable weighted sum, wherein, Indicates the device power level weight, Indicates the device failure influence coefficient, And The weight coefficient of variable And , and satisfy , used to measure And The contribution degree of , by adjusting And Value, can change And The relative importance in determining Value.

[0062] Calculate the data activity index: ,

[0063] Among them, Indicates the calculation result of the Measurement point, The data change amount in unit time, Data update interval, Normalization coefficient.

[0064] Storage area division: according to The storage area is divided into three categories:

[0065] Core area: when And , use high-speed storage medium;

[0066] Ordinary area: when Or ;

[0067] Edge area: when And , use large-capacity storage medium.

[0068] Step 4, time-sequence synchronization conversion between RTU and dispatching master station:

[0069] Timestamp standardization:

[0070] RTU data acquisition time is converted into a second-level timestamp, with the conversion formula being wherein represents a minute value, represents a second value, and has a value range of .

[0071] Storage point serial number calculation:

[0072] Integer serial number calculation:

[0073] The storage point integer serial number is obtained through the formula wherein represents a time value at a certain moment, is usually a time period, is a modulus operator, is the quotient of divided by , and is a time interval, is a rounding function, and is the result of rounding the calculation of , so as to finally obtain the sampling point serial number . Through this formula, the time-related calculation result can be mapped to a discrete sampling point serial number, which is used in data processing or analysis scenarios.

[0074] Non-integer serial number calculation:

[0075] If the non-integer serial number needs to be accurately calculated, the formula can be used to calculate a floating-point value related to time, wherein represents the value at the time point, is a reference time period, represents the result of modulus , that is, the remainder of divided by ; is a time interval parameter. This formula obtains a floating-point value reflecting the relative position of in the period by dividing the remainder of modulus by .

[0076] Step 5, differential data update execution:

[0077] Rule 1: When the RTU data is not refreshed and the storage point serial number is an integer, the data update formula is executed , wherein, represents the data of the th object at the th moment is calculated by weighting the data of the previous moment (the th moment) and the moment before the previous moment (the th moment), the weight coefficient determines the contribution ratio of the previous moment data to the current data, determines the contribution ratio of the moment before the previous moment data, and through this weighting mode, the functions such as smoothing processing or trend prediction of the data are realized.

[0078] Rule 2: If the RTU data is not refreshed and the storage point serial number is not an integer, the data is kept unchanged, and the time delay deviation is recorded.

[0079] Rule 3: When the RTU data is refreshed and the storage point serial number is an integer, the data update formula is executed . Wherein, is the original measurement value, is the system calibration coefficient.

[0080] Rule 4: If the RTU data is refreshed and the storage point serial number is not an integer, different update operations are executed according to the data type:

[0081] Remote signaling quantity: execute ;

[0082] Remote measurement quantity: execute , and trigger the prediction calculation process.

[0083] Measurement data collection and prediction:

[0084] Remote measurement interpolation update:

[0085] The linear interpolation algorithm based on the timestamp is used to update the measurement value of the storage point , and the calculation formula is as follows:

[0086]

[0087] Wherein, and are the absolute moments of the storage points and respectively; is the historical measurement value; is the current RTU refresh value.​​​

[0088] Sliding Window Prediction Model:

[0089] Autoregressive Moving Average Model To predict, this model combines the characteristics of autoregressive models (AR) and moving average models (MA) to capture the dynamic changes in time series data. The model expression is:

[0090] ,

[0091] Where,

[0092] is the autoregressive coefficient, which measures the influence of the sequence value at the past time on the prediction value at the current time; is the moving average coefficient, reflecting the impact of white noise at the past

[0093] time on the current prediction value; and

[0094] , represent the model order of autoregressive and moving average respectively, which determines the number of historical sequence values and historical white noise considered by the model, and needs to be determined according to data characteristics and model fitting effect;

[0095] is the constant term, which can be understood as the long-term mean or benchmark value of the time series;

[0096] represents the white noise sequence, which is a random fluctuation with mean 0 and constant variance, used to describe the random component of the time series that cannot be explained by the autoregressive and moving average parts.

[0097] The model input is the first data in the sliding window, the window size is recommended to be in the range of , by selecting an appropriate window size, the model can balance between capturing sequence trends and adapting to data fluctuations to more accurately predict values.

[0098] Step 6, data collection and prediction:

[0099] Abnormal values are identified by calculating data residuals , where is the predicted value based on historical trends. When the residual exceeds the threshold​​ Time markers are used to identify abnormal data, and the threshold calculation formula is as follows: , The standard deviation of historical residuals. is the confidence coefficient. For marked outlier data, a weighted interpolation method is used for replacement: ,in, and These represent outlier data points. For adjacent normal data points, the replacement value for the abnormal data point is calculated by assigning a weight of 60% to the preceding data point and 40% to the following data point. This smooths the data sequence and reduces the impact of outliers on subsequent analysis.

[0100] Step 7, Initialization and Loop Update:

[0101] A differentiated initialization strategy is executed during system startup:

[0102] when And when there is no historical data, execute on the storage point. ,in This is the first RTU sample value. To initialize the correction coefficients;

[0103] when Furthermore, when historical data exists, a time decay strategy is adopted. , This is the time decay factor.

[0104] like Figure 1 As shown, an RTU data processing device for wide-area measurement of delay includes:

[0105] First processing module: When RTU measurement data is not refreshed and data acquisition time... Corresponding master station data storage point serial number When the value is an integer, it is based on the real-time database set of the scheduling master station. Middle serial number Measurement data of storage points, for data storage point serial numbers The module updates the measurement data. It has an adaptive adjustment mechanism that can dynamically optimize the update algorithm based on the data update frequency and historical data trends, thereby significantly improving the accuracy and timeliness of data updates.

[0106] Second processing module: When the RTU measurement data has not been refreshed and the data acquisition time is... Corresponding master station data storage point serial number In the case of non-integer values, maintain the real-time library collection. The data acquisition time and storage point sequence number remain unchanged. Simultaneously, the module continuously monitors the deviation between the data acquisition time and the storage point sequence number in real time. Once the deviation exceeds a preset threshold, an early warning mechanism is automatically triggered, and a calibration request is sent to the data acquisition terminal to ensure the accuracy of data processing.

[0107] Third processing module: If the RTU measurement data has been refreshed and the data acquisition time... Corresponding master station data storage point serial number If the value is an integer, update the refreshed RTU measurement data to the real-time database set of the scheduling master station. Storage point number Measurement information. This module uses parallel processing technology to achieve synchronous processing of multiple refreshed RTU measurement data, greatly improving data update efficiency and meeting the processing needs of large-scale data.

[0108] Fourth processing module: When RTU measurement data has been refreshed and data acquisition time has expired. Corresponding master station data storage point serial number It is a non-integer, and Real-time database collection with the scheduling master station Storage point number satisfy When dealing with a remote signaling quantity, the updated measurement data will be added to the real-time database of the dispatch master station. The storage point number is The measurement data; if it is telemetry, then update the real-time database set of the scheduling master station. Storage point number of and the storage point sequence number of It performs predictions. This module integrates an intelligent decision-making system that can automatically select the optimal processing strategy based on the type of measurement data (remote signaling or telemetry) and data fluctuation characteristics, ensuring the rationality and scientific nature of data processing.

[0109] The data storage structure construction module is responsible for building a multi-section, time-based, wide-area measurement storage data structure. It stores information about different types of equipment, such as circuit breakers, disconnectors, generators, loads, busbars, capacitors, AC line segments, transformer windings, converters, and DC line segments, through different equipment tables. Each type of equipment uses a two-dimensional data structure to present the equipment's measurement data over a period of time. One column records the equipment ID, and the other column stores the wide-area measurement data set at a fixed time. .gather , Indicates the first A wide-area measurement data, This refers to the storage point number of the wide-area measurement data at the scheduling master station, with a value range of [value range missing]. , This is the sequence number of the largest data storage point. Each wide-area measurement data point... The corresponding scheduling master station real-time database time system has a fixed storage time, the first The value range for each storage time is 0:00:00 to 23:59:59. The maximum retention time for wide-area measurement data is: The fixed time interval between two adjacent wide-area measurement data is , It exceeds the maximum latency of wide-area measurement data. Furthermore, this module employs an intelligent storage strategy based on big data analytics, automatically adjusting the partitioning and layout of data storage according to the real-time operating status of the device and historical data change patterns, further improving data storage and access efficiency.

[0110] Measurement data aggregation module: used to update the real-time database set of the scheduling master station. Storage point number of The calculation formula is (the original text did not provide the formula, so it needs to be added), where , These are the storage point serial numbers. , The corresponding storage time, , These are the storage point serial numbers. , Measurement data, represent RTU measurement data that is constantly updated. Predicting storage point sequence number. of At that time, it will include The former Each measurement data point is input into a pre-defined autoregressive moving average prediction model to calculate the predicted value. When the RTU measurement data is not refreshed and the data acquisition time is... Corresponding master station data storage point serial number When it is 0, it is based on the storage point sequence number. Measurement data update data storage point sequence number Measurement data. In the real-time library collection. During the initialization phase, if the scheduling master station's real-time database set Middle serial number If no measurement data is available, then the current storage point sequence number is... It uses the latest RTU sampling data. This module integrates advanced data quality monitoring and repair mechanisms, utilizes machine learning algorithms to monitor and collect data quality in real time, automatically repairs and corrects abnormal data, and integrates multi-source heterogeneous data through data fusion technology to provide high-quality data support for advanced power system applications.

[0111] While embodiments of the application have been shown and described, it is to be understood that the application is not limited to the details of the embodiments described, since numerous changes, modifications, substitutions and variations can be made thereto without departing from the spirit and scope of the application as defined by the appended claims and their equivalents.

Claims

1. A method for processing RTU data for wide-area measurement delay, characterized in that, Includes the following steps: Step 1: Data storage structure construction: Construct a data storage structure that is compatible with multiple types of power system equipment, establish an independent storage table corresponding to the equipment type, and use a two-dimensional array to classify and store the measurement data. The rows of the two-dimensional array correspond to the equipment identifier and the columns correspond to the storage point sequence number. Step 2, Setting the storage time reference: Set the storage time reference, generate absolute times that correspond one-to-one with the storage point numbers by calculation, and construct a discrete time series to determine the time anchor point of data storage; Step 3: Dual-factor driven storage area partitioning: Based on the dual-factor calculation results of device importance coefficient and data activity index, the storage resources are divided into core storage area, ordinary storage area and edge storage area. The device importance coefficient is determined by combining the device power level and the degree of failure impact, and the data activity index is determined by combining the amount of data change and the update interval. Step 4: Time-Sequence Number Synchronization Conversion between RTU and Scheduling Master Station: Perform time-storage point sequence number conversion, standardize the RTU acquisition time to a second-level timestamp, and calculate the integer or non-integer storage point sequence number through the mapping relationship between the timestamp and the absolute time to achieve time synchronization between the RTU and the scheduling master station. Step 5, Differentiated Data Update Execution: Perform differentiated data updates. Based on the RTU data refresh status and storage point sequence number type, trigger the corresponding update operation. The update operation includes weighted calculation update, keeping the original data unchanged, synchronous overlay update, and classification-triggered prediction update. Step 6, Measurement Data Collection and Prediction: Collect and optimize measurement data, use data prediction algorithms to fill in missing data, identify abnormal data through residual calculation and perform replacement and repair, forming a complete and reliable measurement dataset; Step 7, Initialization and Cyclic Update: Complete system initialization and cyclic update. When the system starts, it will execute the corresponding initialization correction strategy based on historical data. Subsequently, it will continuously execute the operations from Step 4 to Step 6 to achieve dynamic data maintenance.

2. The RTU data processing method and apparatus for wide-area measurement delay according to claim 1, characterized in that, In step 1, the power system equipment is divided into 10 categories: circuit breakers, disconnectors, generators, loads, busbars, capacitors, AC line segments, transformer windings, converters, and DC line segments. Each category of equipment table contains a unique equipment identifier (ID), basic equipment parameters, and a measurement data storage index. The two-dimensional array structure for storing measurement data is defined as follows: The row dimension corresponds to the unique identifier of the device, and the column dimension corresponds to the sequential number of the data storage point of the scheduling master station. Indicates the number of time divisions. To maximize the storage time, It is a fixed time interval.

3. The RTU data processing method and apparatus for wide-area measurement delay according to claim 1, characterized in that, In step 2, the absolute time corresponding to the storage point number is given by the formula. It is confirmed that, among them, At the initial time point, It is a time interval. The sequence number of the time step ( ), Then it represents the first The time values ​​corresponding to each time step are accumulated by multiples of the initial time and the time interval to obtain the values ​​of each time point in the discrete time series.

4. The RTU data processing method and apparatus for wide-area measurement delay according to claim 1, characterized in that, In step 3, the formula for calculating the equipment importance coefficient is as follows: ,in Indicates the weight of the device power level. This represents the impact coefficient of equipment failure. and They are respectively and The weighting coefficients, and ; The formula for calculating the data activity index is as follows: ,in For the first Calculation results for each measurement point The amount of data change per unit time. For the data update interval, This is the normalization coefficient.

5. The RTU data processing method and apparatus for wide-area measurement delay according to claim 1, characterized in that, In step 3, the storage area is divided as follows: Core area: When and At that time, high-speed storage media were used; Normal area: when or hour; Edge region: when and At that time, large-capacity storage media were used.

6. The RTU data processing method and apparatus for wide-area measurement delay according to claim 1, characterized in that, In step 4, timestamp standardization is performed: The conversion formula is ,in, Represents minute values. Represents the value in seconds; Integer sequence number calculation: The formula is ,in The time value at a certain moment. For time period, The modulo operator, For time intervals, This is the rounding function; Calculation of non-integer ordinal numbers: The formula is ,in It is a floating-point value.

7. The RTU data processing method and apparatus for wide-area measurement delay according to claim 1, characterized in that, In step 6, the telemetry interpolation is updated: The formula is ,in and Storage points and The absolute moment; These are historical measurements; This is the current RTU refresh value; The sliding window prediction model uses an autoregressive moving average model. ,in These are the autoregressive coefficients. The moving average coefficient, , These are the model orders for autoregression and moving average, respectively. For constant terms, It is a white noise sequence; Abnormal data handling: residual threshold , The standard deviation of historical residuals. Here is the confidence coefficient, and the outlier replacement formula is: , and These are adjacent normal data points.

8. The RTU data processing method and apparatus for wide-area measurement delay according to claim 1, characterized in that, In step 7, the differentiated initialization strategy is as follows: When there is no historical data, execute ,in This is the first RTU sample value. To initialize the correction coefficients; When historical data exists, a time decay strategy is adopted. , This is the time decay factor.

9. A wide-area measurement delay RTU data processing device, characterized in that, include: First processing module: When the RTU measurement data is not refreshed and the storage point sequence number is an integer, the current storage point data is updated by weighted calculation based on the measurement data of the previous storage point in the real-time database of the scheduling master station, and has an adaptive adjustment mechanism. The second processing module: When the RTU measurement data is not refreshed and the storage point sequence number is not an integer, the real-time library data remains unchanged, the delay deviation is monitored and an early warning and calibration request are triggered. The third processing module: When the RTU measurement data has been refreshed and the storage point sequence number is an integer, the refreshed data will be synchronously updated to the corresponding storage point in the real-time library, using parallel processing technology; Fourth processing module: When the RTU measurement data has been refreshed and the storage point sequence number is not an integer, the remote signaling quantity is directly updated to the corresponding storage point. After the remote measurement is updated, the prediction process is triggered, and the intelligent decision-making system is integrated. Data storage structure construction module: Constructs a multi-section time-wide measurement storage structure, establishes a device table according to device type, uses a two-dimensional data structure to store measurement data, and uses intelligent storage strategies to adjust storage partitions; Measurement data collection module: Updates measurement values ​​through interpolation algorithms, inputs the first N data points in the sliding window into the autoregressive moving average prediction model to calculate the predicted value, and integrates data quality monitoring and repair mechanisms.

Images