A high-voltage equipment multi-source data fusion method based on an Internet of Things card and a collection terminal

Abstract

The application provides a kind of high-voltage equipment multi-source data fusion method based on Internet of Things card and acquisition terminal, comprising: step one, respectively from acquisition terminal and Internet of Things card obtains the sampling density point real-time power data of certain high-voltage equipment in recent N days;Step two, the real-time power data of the same time point of acquisition terminal and Internet of Things card is fused using covariance matrix;Step three, the data after fusion in step two is collected as load curve, and the abnormal data in load curve is overall fitted.The covariance matrix is used to solve the fusion fitting problem of multi-source real-time load data.By fusing multi-source data, the application can effectively solve the problem of poor data quality of Internet of Things card and acquisition terminal, improve the quality and availability of real-time load data, construct load data set, provide data basis for subsequent analysis and fitting, and help promote the close combination of production, teaching and research, and promote the technological innovation and industrial development of power industry.

Description

Technical Field

[0001] This invention relates to the field of real-time load data analysis and processing technology, specifically a method for multi-source data fusion of high-voltage equipment based on IoT cards and acquisition terminals. Background Technology

[0002] During the operation of high-voltage equipment, load data is a crucial indicator reflecting the equipment's operating status and power supply capacity. Real-time monitoring and analysis of high-voltage equipment load data not only provides strong data support for operation and maintenance but also allows for the prediction of future electricity demand through fitting and analysis of historical load data, thereby providing a scientific basis for the planning, design, and operation of the power system.

[0003] However, existing high-voltage equipment suffers from problems such as a lack of uniqueness and reliability in data acquired from multiple sources, as well as low overall quality of real-time load data, which affects business operations and the development of intelligent power systems. Summary of the Invention

[0004] In view of this, the present invention provides a method for multi-source data fusion of high-voltage equipment based on IoT cards and acquisition terminals, which effectively solves the problem of lack of uniqueness and reliability of data acquired when high-voltage equipment has multiple data sources, realizes smooth transition of load data, and improves the smoothness and accuracy of load curves.

[0005] This invention provides a method for multi-source data fusion of high-voltage equipment based on IoT cards and acquisition terminals, comprising the following steps:

[0006] Step 1: Obtain the real-time power data of the sampling density points of a high-voltage device for the most recent N days from the acquisition terminal and the IoT card, respectively.

[0007] Step 2: Use the covariance matrix to fuse the real-time power data of the acquisition terminal and the IoT card at the same time point;

[0008] Step 3: Combine the data from Step 2 into a load curve and perform overall fitting on the abnormal data in the load curve.

[0009] Furthermore, the sampling density is 96 points.

[0010] Furthermore, step one specifically includes:

[0011] Take the active power P of a certain high-voltage equipment obtained from the data acquisition terminal over the past N days at 96 points. AN (P AN 1. P AN 2. P AN 3…P AN 96 ); Take the active power P from 96 points obtained from the IoT card.BN (P BN 1. P BN 2. P BN 3…P BN 96 ).

[0012] Furthermore, step two specifically includes:

[0013] set up:

[0014] P1 AN =(P1) A1 P1 A2 P1 A3 ...P1 AN P2 AN =(P2) A1 P2 A2 P2 A3 ...P2 AN )

[0015] P3 AN =(P3) A1 P3 A2 P3 A3 ...P3 AN ...P 96 AN =(P 96 A1 P 96 A2 P 96 A3 ...P 96 AN )

[0016] P is obtained from the covariance matrix. n AN The data distribution trend, where n = 1 to 96, yields the 96-point covariance matrix ∑ of a high-voltage device acquired from the acquisition terminal over the most recent N days. n AN The formula is as follows:

[0017]

[0018] Similarly, let's set:

[0019] P1 BN =(P1) B1 P1 B2 P1 B3 ...P1 BN P2 BN =(P2) B1 P2 B2 P2 B3 ...P2BN )

[0020] P3 BN =(P3) B1 P3 B2 P3 B3 ...P3 BN ...P 96 BN =(P 96 B1 P 96 B2 P 96 B3 ...P 96 BN )

[0021] P is obtained from the covariance matrix. n BN The data distribution trend was used to derive the 96-point covariance matrix ∑ of a high-voltage device acquired from the acquisition terminal over the past N days. n BN The formula is as follows:

[0022]

[0023] The CIF algorithm was used to analyze ∑ n AN and ∑ n BN The covariance correlation coefficient between the two was obtained by performing a fusion calculation. The calculation formula is as follows:

[0024] Where ∑ is the corresponding covariance matrix;

[0025] Based on the ∑ obtained above, it is substituted as a weighting coefficient into the calculation to obtain the active power at each time point. The calculation formula is as follows:

[0026] Where P is the state variable, i.e., the active power at each time point n).

[0027] Furthermore, step three involves performing an overall fitting of the abnormal data in the load curve, specifically including:

[0028] (1) Abnormal data identification

[0029] Curve anomalies are categorized into sudden increases, negative values, and null values. The order of judgment is null values, negative values, and sudden increases. Once an anomaly at a certain point in time is assigned to a specific anomaly in the above order, other anomalies are no longer calculated.

[0030] (2) Abnormal data processing

[0031] When a sudden increase occurs, the data from the previous time point is used as the data for the current time point, i.e., P. n =P n-1 Once the data is returned, the returned values ​​will be used to supplement it.

[0032] When a negative value occurs, the current time point is set to zero, i.e., P. n =0;

[0033] When a null value is encountered, the current time point is set to zero, i.e., P. n =0, and the returned value will be used to fill in the gaps after the data is returned.

[0034] Furthermore, the specific rules for identifying sudden increases, negative values, and null values ​​in the abnormal data identification process are as follows:

[0035] Sudden increase: P n Greater than twice the maximum power of the electricity meter, that is Where Ue is the rated voltage, I max Where K is the maximum current and K is the overload factor;

[0036] Negative value: P n <0.

[0037] Null value: P n =null.

[0038] Compared with existing technologies, this invention proposes a multi-source data fusion method for high-voltage equipment based on IoT cards and acquisition terminals. By fusing multi-source data, it can effectively solve the problem of poor data quality from IoT cards and acquisition terminals, improve the quality and availability of real-time load data, construct a load dataset, provide a data foundation for subsequent analysis and fitting, and help promote close integration of industry, academia and research, and drive technological innovation and industrial development in the power industry. Attached Figure Description

[0039] Figure 1 This is a flowchart of a multi-source data fusion method for high-voltage equipment based on an IoT card and a data acquisition terminal, according to an embodiment of the present invention. Detailed Implementation

[0040] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0041] This invention provides a method for multi-source data fusion of high-voltage equipment based on IoT cards and acquisition terminals. The implementation process is as follows: Figure 1 As shown, the main steps are as follows: (1) Obtain real-time load data from the high-voltage equipment meters at 15-minute intervals from the IoT card and the acquisition terminal respectively; (2) Use the covariance matrix to fuse the two types of data at the same time point; (3) Compile the fused data into a 96-point load curve to complete the fitting of the overall load curve. The specific implementation process is as follows:

[0042] Step 1: Obtain 96 active power points P from the data acquisition terminal for a certain high-voltage equipment over the past N days. AN (P AN 1. P AN 2. P AN 3…P AN 96 ); Take the active power P from 96 points obtained from the IoT card. BN (P BN 1. P BN 2. P BN 3…P BN 96 ).

[0043] Step two, assuming

[0044] P1 AN =(P1) A1 P1 A2 P1 A3 ...P1 AN P2 AN =(P2) A1 P2 A2 P2 A3 ...P2 AN )

[0045] P3 AN =(P3) A1 P3 A2 P3 A3 ...P3 AN ...P 96 AN =(P 96 A1 P 96 A2 P 96 A3 ...P 96 AN )

[0046] P is obtained from the covariance matrix. n AN The data distribution trend, where n = 1 to 96, yields the 96-point covariance matrix ∑ of a high-voltage device acquired from the acquisition terminal over the most recent N days. n ANThe formula is as follows:

[0047]

[0048] Similarly, let's set:

[0049] P1 BN =(P1) B1 P1 B2 P1 B3 ...P1 BN P2 BN =(P2) B1 P2 B2 P2 B3 ...P2 BN )

[0050] P3 BN =(P3) B1 P3 B2 P3 B3 ...P3 BN )P 96 BN =(P 96 B1 P 96 B2 P 96 B3 ...P 96 BN ) ...

[0052] P is obtained from the covariance matrix. n BN The data distribution trend was used to derive the 96-point covariance matrix ∑ of a high-voltage device acquired from the acquisition terminal over the past N days. n BN The formula is as follows:

[0053]

[0054] Through CIF for ∑ n AN and ∑ n BN The covariance correlation coefficient between the two was obtained by performing a fusion calculation. The calculation formula is as follows:

[0055] Where ∑ is the corresponding covariance matrix;

[0056] Based on the obtained ∑, it is substituted as a weighting coefficient into the calculation to obtain the active power at each time point. The calculation formula is as follows:

[0057] Where P is the state variable, that is, the active power at each time point n (n = 1 to 96).

[0058] Step 3: Based on the 96 real-time load data points of the high-voltage equipment generated in the above two steps, a load curve is formed. Then, the abnormal data in the load curve is fitted as a whole. The specific implementation steps are as follows:

[0059] (1) Abnormal data identification

[0060] Curve anomalies are categorized into sudden increases, negative values, and null values. The order of judgment is null values, negative values, and sudden increases.

[0061] Once an anomaly at a certain point in time is assigned to a specific anomaly in the order described above, other anomalies will not be considered.

[0062] Sudden increase: P n Greater than twice the maximum power of the electricity meter, that is Where Ue is the rated voltage, I max K is the maximum current, and K is the overload factor (based on voltage overvoltage of 1.2 times and current overload of 1.2 times, the overload factor is appropriately increased, and is temporarily set to 2).

[0063] Negative value: P n <0.

[0064] Null value: P n =null.

[0065] (2) Abnormal data handling methods

[0066] When a sudden increase occurs, the data from the previous time point is used as the data for the current time point, i.e., P. n =P n-1 Once the data is returned, the returned values ​​will be used to supplement it.

[0067] When a negative value occurs, the current time point is set to zero, i.e., P. n =0.

[0068] When a null value is encountered, the current time point is set to zero, i.e., P. n =0, and the returned value will be used to fill in the gaps after the data is returned.

[0069] As can be seen from the above, in order to solve the problem of multi-source data fusion for high-voltage equipment, this invention proposes a multi-source data fusion method for high-voltage equipment based on IoT cards and acquisition terminals. First, it is based on 15-minute high-frequency data acquisition from IoT cards and acquisition terminals. Then, it uses the covariance matrix to fuse the multi-source data. Finally, it performs overall data fitting on the fused data to improve the load data quality of high-voltage equipment in real-time monitoring. It has high reliability and uniqueness in handling abnormal load curve data of individual high-voltage equipment.

[0070] This invention primarily addresses the problem of how to fuse multi-source data to fit a complete 96-point (or 288-point, 1440-point) load curve when high-voltage equipment has multiple data sources. By using a covariance matrix, real-time power data acquired from IoT cards and acquisition terminals at each time point are fused. The fused single-time-point data undergoes difference analysis to determine if further fitting is needed. If further fitting is required, the load data at the current time point is fitted using preceding and following time points. This effectively solves the problem of lack of uniqueness and reliability in data acquired from multiple sources in high-voltage equipment, achieving a smooth transition of load data and improving the smoothness and accuracy of the load curve.

[0071] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for multi-source data fusion of high-voltage equipment based on IoT cards and acquisition terminals, characterized in that, Includes the following steps: Step 1: Obtain the real-time power data of the sampling density points of a high-voltage device for the most recent N days from the acquisition terminal and the IoT card, respectively. Step 2: Use the covariance matrix to fuse the real-time power data of the acquisition terminal and the IoT card at the same time point; Step 3: Combine the data from Step 2 into a load curve and perform overall fitting on the abnormal data in the load curve; Step one specifically includes: Take the active power P of a certain high-voltage equipment obtained from the data acquisition terminal over the past N days at 96 points. AN (P) AN 1. P AN 2. P AN 3…P AN 96 ); Take the active power P from 96 points obtained from the IoT card BN (P) BN 1. P BN 2. P BN 3…P BN 96 ); Step two specifically includes: set up: ； ... ； P is obtained from the covariance matrix. n AN The data distribution trend, where n=1~96, yields the 96-point covariance matrix ∑ of a high-voltage device acquired from the acquisition terminal in the most recent N days. n AN The formula is as follows: ； Similarly, let's set: ； ... ； P is obtained from the covariance matrix. n BN The data distribution trend was used to derive the 96-point covariance matrix ∑ of a high-voltage device acquired from the acquisition terminal over the past N days. n BN The formula is as follows: ； The CIF algorithm was used to analyze ∑ n AN and ∑ n BN The covariance correlation coefficient between the two was obtained by performing a fusion calculation. The calculation formula is as follows: , where ∑ is the corresponding covariance matrix; Based on the ∑ obtained above, it is substituted as a weighting coefficient into the calculation to obtain the active power at each time point. The calculation formula is as follows: , where P is the state variable, i.e. the active power at each time point n.

2. The method for multi-source data fusion of high-voltage equipment based on IoT cards and acquisition terminals as described in claim 1, characterized in that: The sampling device must collect samples at the same time point, regardless of the density and average of the collected samples.

3. The method for multi-source data fusion of high-voltage equipment based on IoT cards and acquisition terminals as described in claim 1, characterized in that: Step three involves performing an overall fitting of the outlier data in the load curve, specifically including: (1) Judgment of abnormal data Curve anomalies are categorized into sudden increases, negative values, and null values. The order of judgment is null values, negative values, and sudden increases. Once an anomaly at a certain point in time is assigned to a specific anomaly in the above order, other anomalies are no longer calculated. (2) Abnormal data processing When a sudden increase occurs, the data from the previous time point is used as the data for the current time point. Once the data is returned, the returned values ​​will be used to supplement it. When a negative value occurs, the current time point is set to zero. ; When a null value is encountered, the current time point is set to zero. Once the data is returned, the returned values ​​will be used to supplement it.

4. The method for multi-source data fusion of high-voltage equipment based on IoT cards and acquisition terminals as described in claim 3, characterized in that: The specific rules for identifying sudden increases, negative values, and null values ​​in the anomaly data identification process are as follows: Sudden increase: P n Greater than twice the maximum power of the electricity meter, that is Where Ue is the rated voltage, I max Where K is the maximum current and K is the overload factor; Negative values: ; Null value: .

Images