Transverse metering tracing number-setting interpolation algorithm based on source-load aggregation application

Abstract

The invention discloses a traversal measurement tracing number-setting interpolation algorithm based on source-load aggregation application, which comprises the following steps of: S1, selecting three points before and after missing data as original data, and learning an interpolation model to fit the original data through a cubic polynomial; and obtaining a cubic polynomial function coefficient through a least square method, and finally substituting the missing data into a cubic function to complete interpolation of the missing data. According to the traversal metering tracing number-setting interpolation algorithm based on the source load aggregation application, adjacent three-point data is selected to carry out cubic function interpolation of the least square method, and context information around missing points is fully considered, so that the trend and characteristics of original data are better restored, the average mean square error reaches the magnitude order of 10-3 power, and the calculation accuracy is improved. The missing load data is well interpolated, the deviation between the interpolation result and the real data is relatively small, and the validity of the interpolation algorithm in the aspect of interpolation accuracy is embodied.

Description

Technical Field

[0001] The present invention belongs to the technical field of load data interpolation, and in particular relates to a cross-metering tracing and data interpolation algorithm based on source-load aggregation application. Background Art

[0002] The large-scale access of distributed power sources and electric vehicles to the power grid has led to complex waveforms and power flows at gateway metering points. In addition, gateway metering data presents a complex structure, and the data collection of power consumption information of dedicated transformer users faces the difficulties of unclear historical archives, complex data sources, and chaotic topology and flow, which makes fusion analysis difficult.

[0003] Therefore, further improvements are made to the above problems. Summary of the invention

[0004] The main purpose of the present invention is to provide a crossing metering tracing and interpolation algorithm based on source-load aggregation application, select three adjacent point data for cubic function interpolation using the least squares method, and fully consider the contextual information around the missing point, so as to better restore the trend and characteristics of the original data. The average mean square error reaches the order of 10 to the -3 power, and the missing load data is interpolated well. The deviation between the interpolation result and the real data is relatively small, which reflects the effectiveness of this interpolation algorithm in terms of interpolation accuracy.

[0005] To achieve the above objectives, the present invention discloses a cross-metering tracking and interpolation algorithm based on source-load aggregation application, comprising the following steps:

[0006] Step S1: Select three points before and after the missing data (data to be interpolated) as the original data, and the interpolation model learns a cubic polynomial to fit the original data, that is:

[0007] f(x)=p3x 3 +p2x 2 +p1x+p0;

[0008] The cubic polynomial function coefficients are obtained by the least squares method, and finally the missing data are brought into the cubic function to complete the interpolation of the missing data, where:

[0009] The least squares method solves the unknown polynomial coefficients p3, p2, p1, and p0 in the model based on the minimization of the mean square error:

[0010]

[0011]

[0012] E is respectively i Taking partial derivatives, we get the following formula:

[0013]

[0014] Let the partial derivative equal to zero and simplify to get the following formula:

[0015]

[0016] Among them, E: minimum sum of squares of deviations; P3: undetermined coefficient column vector 3; P2: undetermined coefficient column vector 2; P1: undetermined coefficient column vector 2; P0: undetermined coefficient column vector 0; i: infinite growth value; X: coordinate X axis; Y: coordinate Y axis; xi: infinite growth vector on coordinate X axis; yi: infinite growth vector on coordinate Y axis;

[0017] Transformed into a matrix expression as follows, the problem is transformed into a four-variable linear equation system problem:

[0018] X T Xp=X T Y;

[0019] in,

[0020] Then the unknown coefficient matrix is ​​solved: p = (X T X) -1 X T Y.

[0021] As a further preferred technical solution of the above technical solution, in step S1:

[0022] Select the time and load data of three points before and after the missing data as the original data, and obtain the input data time data x1, x2, ..., x6, and the corresponding load data y1, y2, ..., y6;

[0023] Calculating the Matrix and

[0024] Calculate the matrix X transpose X respectively T , matrix multiplication X T X and X T Y, matrix inversion (X T X) -1 ;

[0025] According to the formula p=(X T X) -1 X T Y, matrix multiplication to obtain the final column vector of undetermined coefficients

[0026] According to the formula f(x) = p3x 3 +p2x 2+p1x+p0 calculates the load data at the missing time and interpolates the missing load data.

[0027] As a further preferred technical solution of the above technical solution, the collected one-dimensional electric energy data sequence is reorganized with the one-dimensional electric energy data sequence collected in the same period of the previous year to obtain the one-dimensional electric energy data sequence to be analyzed, and then converted into a matrix with the same anti-diagonal element values, and then a fitting matrix is ​​established, a decomposition objective function is established, and the local optimal solution of the objective function is calculated using the stochastic gradient descent method, and finally the missing data is estimated through the corresponding elements of the fitting matrix.

[0028] To achieve the above objectives, the present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the steps of the cross-metering tracking and interpolation algorithm based on source-load aggregation application are implemented.

[0029] To achieve the above objectives, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the cross-metering tracking and interpolation algorithm based on source-load aggregation application. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 It is a flow chart of the cross-metering tracking and setting interpolation algorithm based on source-load aggregation application of the present invention.

[0031] Figure 2 It is a schematic diagram of the interpolation results of the cross-metering tracking and setting interpolation algorithm based on source-load aggregation application of the present invention. DETAILED DESCRIPTION

[0032] The following description is used to disclose the present invention so that those skilled in the art can implement the present invention. The preferred embodiments described below are only examples, and those skilled in the art can think of other obvious variations. The basic principles of the present invention defined in the following description can be applied to other embodiments, variations, improvements, equivalents, and other technical solutions that do not deviate from the spirit and scope of the present invention.

[0033] In the preferred embodiments of the present invention, those skilled in the art should note that the original data and the like involved in the present invention may be regarded as prior art.

[0034] Preferred embodiments.

[0035] The present invention discloses a crossing metering tracking and setting interpolation algorithm based on source-load aggregation application, which is characterized by comprising the following steps:

[0036] Step S1: Select three points before and after the missing data (data to be interpolated) as the original data, and the interpolation model learns a cubic polynomial to fit the original data, that is:

[0037] f(x)=p3x 3 +p2x 2 +p1x+p0;

[0038] The cubic polynomial function coefficients are obtained by the least squares method, and finally the missing data are brought into the cubic function to complete the interpolation of the missing data, where:

[0039] The least squares method solves the unknown polynomial coefficients p3, p2, p1, and p0 in the model based on the minimization of the mean square error:

[0040]

[0041]

[0042] E is respectively i Taking partial derivatives, we get the following formula:

[0043]

[0044] Let the partial derivative equal to zero and simplify to get the following formula:

[0045]

[0046] Among them, E: minimum sum of squares of deviations; P3: undetermined coefficient column vector 3; P2: undetermined coefficient column vector 2; P1: undetermined coefficient column vector 2; P0: undetermined coefficient column vector 0; i: infinite growth value; X: coordinate X axis; Y: coordinate Y axis; xi: infinite growth vector on coordinate X axis; yi: infinite growth vector on coordinate Y axis;

[0047] Transformed into a matrix expression as follows, the problem is transformed into a four-variable linear equation system problem:

[0048] X T Xp=X T Y;

[0049] in,

[0050] Then the unknown coefficient matrix is ​​solved: p = (X T X) -1 X T Y.

[0051] Specifically, in step S1:

[0052] Select the time and load data of three points before and after the missing data as the original data, and obtain the input data time data x1, x2, ..., x6, and the corresponding load data y1, y2, ..., y6;

[0053] Calculating the Matrix and

[0054] Calculate the matrix X transpose X respectively T , matrix multiplication X T X and X T Y, matrix inversion (X T X) -1 ;

[0055] According to the formula p=(X T X) -1 X T Y, matrix multiplication to obtain the final column vector of undetermined coefficients

[0056] According to the formula f(x) = p3x 3 +p2x 2 +p1x+p0 calculates the load data at the missing time and interpolates the missing load data.

[0057] More specifically, the collected one-dimensional electric energy data sequence is reorganized with the one-dimensional electric energy data sequence collected in the same period of the previous year to obtain the one-dimensional electric energy data sequence to be analyzed, and then converted into a matrix with the same anti-diagonal element values, and then a fitting matrix is ​​established, a decomposition objective function is established, and the stochastic gradient descent method is used to calculate the local optimal solution of the objective function, and finally the missing data is estimated through the corresponding elements of the fitting matrix.

[0058] The present invention also discloses an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, the steps of the cross-metering tracking and interpolation algorithm based on source-load aggregation application are implemented.

[0059] The present invention also discloses a non-transitory computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the steps of the crossing metering tracking and setting interpolation algorithm based on source-load aggregation application are implemented.

[0060] For the present invention:

[0061] Select three points before and after the missing data (data to be interpolated) as the original data, and fit the original data through the cubic function. Through the least squares method, the cubic function coefficients p3, p2, p1, and p0 in the model are solved based on the minimization of the mean square error, and finally the missing data are brought into the cubic function to complete the interpolation of the missing data. The collected one-dimensional electric energy data sequence is reorganized with the one-dimensional electric energy data sequence collected in the same period of the previous year to obtain the one-dimensional electric energy data sequence to be analyzed, and then converted into a matrix with the same anti-diagonal element value, and then a fitting matrix is ​​established, a decomposition objective function is established, and the local optimal solution of the objective function is calculated by the stochastic gradient descent method. Finally, the missing data is estimated by the corresponding elements of the fitting matrix.

[0062] The missing data filling method of gateway electricity metering that integrates historical electricity consumption data uses historical electricity data of the same period and current electricity data for processing, which can greatly reduce the time cost of model parameter updating and filling. It uses data from historical moments to calculate missing data, introduces the seasonal attributes of electricity load, and improves the reliability of filled data.

[0063] It is worth mentioning that the technical features such as raw data involved in the patent application of this invention should be regarded as the prior art. The specific structure, working principle and possible control method and spatial layout method of these technical features can be selected by conventional methods in the field and should not be regarded as the inventive point of this patent. This patent will not be further elaborated.

[0064] For those skilled in the art, it is still possible to modify the technical solutions described in the aforementioned embodiments, or to make equivalent substitutions for some of the technical features therein. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A crossing metering tracking and interpolation algorithm based on source-load aggregation application, characterized in that: The following steps are involved: Step S1: Select three points before and after the missing data as the original data, and the interpolation model learns a cubic polynomial to fit the original data, that is: f(x)=p3x 3 +p2x 2 +p1x+p0; The cubic polynomial function coefficients are obtained by the least squares method, and finally the missing data are brought into the cubic function to complete the interpolation of the missing data, where: The least squares method solves the unknown polynomial coefficients p3, p2, p1, and p0 in the model based on the minimization of the mean square error: E is respectively i Taking partial derivatives, we get the following formula: Let the partial derivative equal to zero and simplify to get the following formula: Among them, E: minimum sum of squares of deviations; P3: undetermined coefficient column vector 3; P2: undetermined coefficient column vector 2; P1: undetermined coefficient column vector 2; P0: undetermined coefficient column vector 0; i: infinite growth value; X: coordinate X axis; Y: coordinate Y axis; xi: infinite growth vector on coordinate X axis; yi: infinite growth vector on coordinate Y axis; Transformed into a matrix expression as follows, the problem is transformed into a four-variable linear equation system problem: X T Xp=X T AND; in, Then the unknown coefficient matrix is ​​solved: p = (X T X) -1 X T Y.

2. According to the method of claim 1, the method is characterized in that: In step S1: Select the time and load data of three points before and after the missing data as the original data, and obtain the input data time data x1, x2, ..., x6, and the corresponding load data y1, y2, ..., y6; Calculating the Matrix and Calculate the matrix X transpose X respectively T , matrix multiplication X T X and X T Y, matrix inversion (X T X) -1 ; According to the formula p=(X T X) -1 X T Y, matrix multiplication to obtain the final column vector of undetermined coefficients According to the formula f(x) = p3x 3 +p2x 2 +x1x+p0 calculates the load data at the missing time and interpolates the missing load data.

3. According to the method of claim 2, the algorithm for traversal measurement tracking and interpolation based on source-load aggregation application is characterized in that: The collected one-dimensional electric energy data sequence is reorganized with the one-dimensional electric energy data sequence collected in the same period of the previous year to obtain the one-dimensional electric energy data sequence to be analyzed, and then it is converted into a matrix with the same anti-diagonal element values. Then, a fitting matrix is ​​constructed, a decomposition objective function is established, and the local optimal solution of the objective function is calculated using the stochastic gradient descent method. Finally, the missing data is estimated through the corresponding elements of the fitting matrix.

4. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the steps of the crossing metering tracing and interpolation algorithm based on source-load aggregation application are implemented as described in any one of claims 1 to 3.

5. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the crossing metering tracking and interpolation algorithm based on source-load aggregation application as described in any one of claims 1 to 3 are implemented.

Images