A smart meter error perception evaluation method and system based on big data

Abstract

A smart meter error perception evaluation method and system based on big data. The method first creates three empty sets Z A , Z B , Z C , and obtain the basic error matrix based on the first inspection and operation test data of the smart meter, and then calculate the average error matrix #imgabs0# variance matrix S, perception coefficient a in turn based on the basic error matrix of the smart meter ij , perception matrix mean #imgabs1 #perception matrix variance S′ j , and then determine whether #imgabs2# is satisfied. If so, then at #imgabs3# the load error X of the smart meter is set j Merge into set Z A , in a ij ∈S, X j Merge into set Z B , in a ij ∈V, X j Merge into set Z C , then determine whether i>k and j>q are satisfied. If so, output the set Z A , Z B , Z C Otherwise, proceed to the next perception cycle, and finally based on X j and set Z A , Z B , Z C The invention can accurately screen out the meters with metering risks in the smart grid and effectively reduce the failure risk of the smart meters during use.

Description

Technical Field

[0001] The present invention belongs to the field of electric meter error perception, and specifically relates to a smart electric meter error perception evaluation method and system based on big data. Background Art

[0002] In the context of large-scale electric vehicles connected to the power grid, the accuracy of smart meter energy measurement is particularly important. Smart meters are the data basis for computer measurement, auditing and analysis, and are closely related to our daily lives. With the rapid development of power technology, my country has installed hundreds of millions of smart meters, and the number of basic electronic components will reach tens of billions. Even if manufacturers can control the defective rate below 3.4 parts per million according to the 6σ technology, there are still a large number of meters with component defects. Instruments with defective components have a greater probability of failure than other instruments. For example, due to a certain deviation between the actual value of the resistor and the nominal value, or the influence of external factors such as voltage and current, the value of the instrument will drift linearly or nonlinearly. Such problems include man-made damage during transportation, which may not be manifested during the initial inspection; but in subsequent use, it is more likely to have error drift or failure, resulting in power monitoring and statistical errors. Therefore, the perception and evaluation of smart meter errors are very important. Summary of the invention

[0003] The purpose of the present invention is to provide a smart meter error perception evaluation method and system based on big data in order to address the above-mentioned problems existing in the prior art.

[0004] To achieve the above objectives, the technical solution of the present invention is as follows:

[0005] In a first aspect, the present invention proposes a smart meter error perception evaluation method based on big data, comprising:

[0006] S1. Create three empty sets Z A , Z B , Z C , and obtain the basic error matrix V = e based on the first inspection and operation test data of the smart meter ij , where i = 1, 2, ..., m, j = 1, 2, ..., n, m is the number of smart meters, and n is the number of load points;

[0007] S2, calculate the average error matrix based on the basic error matrix V Variance matrix S, perception coefficient a ij , the average value of the perception matrix Perception matrix variance S j ′;

[0008] S3. Determine whether If satisfied, go to S4, otherwise go to S5;

[0009] S4. If Then the load error of the smart meter is j Merge into set Z A If a ij ∈S, then X j Merge into set Z B If a ij ∈V, then X j Merge into set Z C ;

[0010] S5. Determine whether i>k and j>q. If so, output set Z. A , Z B , Z C , otherwise return to S3 for the next perception cycle, where k = m, q = n + 2;

[0011] S6, based on X j and set Z A , Z B , Z C Determine the risk level of smart meters.

[0012] Said S is calculated based on the following formula:

[0013]

[0014]

[0015]

[0016]

[0017] In the above formula, is the average error of the i-th smart meter, S i is the variance of the i-th smart meter.

[0018] The a ij , S j ' is calculated based on the following formula:

[0019]

[0020]

[0021]

[0022] In the above formula, V′ k,q is the perception matrix, a ij is the perception coefficient of the i-th smart meter to the j-th load point.

[0023] In S6, the risk level determination of the smart meter includes:

[0024] If X j Not in Z A , Z B , Z C If the smart meter is in the union of , it is judged as a low-risk smart meter;

[0025] If X j In the collection Z A If the value is medium, it is judged as a medium-risk smart meter;

[0026] If X j In Z B , Z C If the number of meters is in the same concentration, it is judged as a high-risk smart meter;

[0027] If X j In Z B , Z C If it is in the intersection of , it is determined to be a very high-risk smart meter.

[0028] In a second aspect, the present invention proposes a smart meter error perception assessment system based on big data, comprising a set creation module, a perception calculation module, a first judgment module, a set update module, a second judgment gate module, and a risk level determination module;

[0029] The set creation module is used to create three empty sets Z A , Z B , Z C , and obtain the basic error matrix V = e based on the first inspection and operation test data of the smart meter ij , where i = 1, 2, ..., m, j = 1, 2, ..., n, m is the number of smart meters, and n is the number of load points;

[0030] The perception calculation module is used to calculate the average error matrix based on the basic error matrix V in turn. Variance matrix S, perception coefficient a ij , the average value of the perception matrix Perception matrix variance S j ′;

[0031] The first judgment module is used to judge whether If satisfied, the set update module is activated, otherwise the second judgment gate module is activated;

[0032] The collection updating module is used to The load error of the smart meter is X j Merge into set Z A , in a ij ∈S, when Xj Merge into set Z B , in a ij ∈V, X j Merge into set Z C ;

[0033] The second judgment gate module is used to judge whether i>k and j>q are satisfied. If so, the output set Z is A , Z B , Z C , otherwise activate the first judgment module to perform the next perception cycle, where k=m, q=n+2;

[0034] The risk level determination module is used to determine the risk level based on X j and set Z A , Z B , Z C Determine the risk level of smart meters.

[0035] The perception calculation module is calculated based on the following formula S:

[0036]

[0037]

[0038]

[0039]

[0040] In the above formula, is the average error of the i-th smart meter, S i is the variance of the i-th smart meter.

[0041] The perception calculation module calculates a based on the following formula ij , S j ′:

[0042]

[0043]

[0044]

[0045] In the above formula, V′ k,q is the perception matrix, a ij is the perception coefficient of the i-th smart meter to the j-th load point.

[0046] The risk level determination module determines the risk level of the smart meter based on the following method:

[0047] If X j Not in Z A , Z B , Z C If the smart meter is in the union of , it is judged as a low-risk smart meter;

[0048] If X j In the collection Z A If the value is medium, it is judged as a medium-risk smart meter;

[0049] If X j In Z B , Z C If the number of meters is in the same concentration, it is judged as a high-risk smart meter;

[0050] If X j In Z B , Z C If the smart meter is in the intersection of , it is judged as a very high-risk smart meter.

[0051] Compared with the prior art, the present invention has the following beneficial effects:

[0052] The present invention proposes a smart meter error perception evaluation method based on big data. First, three empty sets Z are created. A , Z B , Z C , and obtain the basic error matrix based on the first inspection and operation test data of the smart meter, and then calculate the average error matrix based on the basic error matrix of the smart meter Variance matrix S, perception coefficient a ij , the average value of the perception matrix Perception matrix variance S j ′, and then judge whether it satisfies If satisfied, then The load error of the smart meter is X j Merge into set Z A , in a ij ∈S, X j Merge into set Z B , in a ij ∈V, X j Merge into set Z C , then determine whether i>k and j>q are satisfied. If so, output the set Z A , Z B , Z C Otherwise, proceed to the next perception cycle, and finally based on X j and set Z A , Z B , Z CTo determine the risk level of smart meters, this method adopts an error-aware cyclic screening algorithm, which can classify the metering risks of smart meters, accurately screen out meters with metering risks in the smart grid, and effectively reduce the failure risk of smart meters during use. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 This is a flowchart of the perception cycle in Example 1.

[0054] Figure 2 This is a rule diagram for determining the risk level in Example 1.

[0055] Figure 3 This is a structural diagram of the system described in Example 2. DETAILED DESCRIPTION

[0056] The present invention is further described in detail below in conjunction with specific implementations and drawings.

[0057] The overall performance error of a normal component should be linear, and defective components exhibit three main problems: first, the component itself has an inherent error, but the overall performance is still linear; second, the component error is generally nonlinear; third, the component has a large or small error at a certain point (or range), while it remains linear in other ranges. The first of the three defect types will not substantially affect the service life and has a low failure rate, while the second and third defects greatly increase the failure of the device, resulting in a low service life or a high failure rate.

[0058] The present invention divides the screening algorithms of smart meters into three categories: A, B, and C, each of which corresponds to one of the above-mentioned faults:

[0059] The A-type algorithm determines the serious error x in the same column of the basic error matrix V. j (j=1,2,3…n), it can screen out meters whose errors at the same load point are significantly different from those at other points. The Class A algorithm screens out problems where the components themselves have fixed errors, but the overall performance is still linear.

[0060] The Class B algorithm can determine whether the load point error of a single instrument is seriously deviated from the load points of other instruments by calculating the average value of each row in the basic error matrix V. The Class B algorithm screens out problems with error nonlinearity.

[0061] The C algorithm calculates the standard deviation S of each row in the basic error matrix V j (j=1,2,3…n), it can be determined whether the error of all load points of the same meter in the same batch is offset as a whole compared with other meters. The Class C algorithm screens out problems where the error is very large or very small at a certain point (or range) while maintaining linearity in other ranges.

[0062] The above three types of screening algorithms play an important role in the entire system. Three different screening factors are used to analyze the verification error of the electric energy meter. Three different screening factors are used to determine the defect type of the component. The electric energy meters with metering risks can be accurately screened out from a large number of electric energy meters.

[0063] For smart meters that are judged to be high-risk, they will be tracked and monitored during subsequent use; for smart meters that are judged to be extremely high-risk, they will be monitored and controlled during subsequent use, and on-site calibration will be carried out on schedule. By determining the metering risk level of each smart meter and listing the meters that need to be monitored, it is convenient to track and monitor during use.

[0064] Embodiment 1:

[0065] like Figure 1 As shown, a smart meter error perception evaluation method based on big data is implemented in the following steps:

[0066] 1. Create three empty sets Z A , Z B , Z C , corresponding to the smart meter error problems perceived by the three algorithms A, B, and C, and the basic error matrix V = e is obtained based on the first inspection and operation test data of the smart meter. ij , where i = 1, 2, ..., m, j = 1, 2, ..., n, m is the number of smart meters, and n is the number of load points.

[0067] In this embodiment, each smart meter has 10 load points, and the error matrix V constructed is as follows:

[0068]

[0069] 2. Calculate the average error matrix based on the basic error matrix of the smart meter Variance matrix S, perception coefficient a ij , the average value of the perception matrix Perception matrix variance S j ′, the calculation formula is as follows:

[0070]

[0071]

[0072]

[0073]

[0074]

[0075]

[0076]

[0077] In the above formula, is the average error of the i-th smart meter, S i is the variance of the ith smart meter, V′ k,q is the perception matrix, a ij is the perception coefficient of the i-th smart meter to the j-th load point.

[0078] 3. Determine whether it is satisfied If satisfied, go to step 4, otherwise go to step 5.

[0079] 4. If Then the load error of the smart meter is j Merge into set Z A If a ij ∈S, then X j Merge into set Z B If a ij ∈V, then X j Merge into set Z C .

[0080] 5. Determine whether i>k and j>q are satisfied. If so, output set Z A , Z B , Z C , otherwise return to S3 for the next perception cycle, where k=m, q=n+2.

[0081] 6. Based on X j and set Z A , Z B , Z C Determine the risk level of smart meters, such as Figure 2 As shown, including:

[0082] If X j Not in Z A , Z B , Z C If the smart meter is in the union of , it is judged as a low-risk smart meter;

[0083] If X j In the collection Z A If the value is medium, it is judged as a medium-risk smart meter;

[0084] If X j In Z B , Z C If the number of meters is in the same concentration, it is judged as a high-risk smart meter;

[0085] If X jIn Z B , Z C If it is in the intersection of , it is determined to be a very high-risk smart meter.

[0086] Embodiment 2:

[0087] like Figure 3 As shown, a smart meter error perception and evaluation system based on big data includes a set creation module, a perception calculation module, a first judgment module, a set update module, a second judgment gate module, and a risk level determination module.

[0088] The set creation module is used to create three empty sets Z A , Z B , Z C , and obtain the basic error matrix V = e based on the first inspection and operation test data of the smart meter ij , where i = 1, 2, ..., m, j = 1, 2, ..., n, m is the number of smart meters, and n is the number of load points;

[0089] The perception calculation module is used to calculate the average error matrix based on the basic error matrix V in turn. Variance matrix S, perception coefficient a ij , the average value of the perception matrix Perception matrix variance S j ′, the specific calculation formula is as follows:

[0090]

[0091]

[0092]

[0093]

[0094]

[0095]

[0096]

[0097] In the above formula, is the average error of the i-th smart meter, S i is the variance of the ith smart meter, V′ k,q is the perception matrix, a ij is the perception coefficient of the i-th smart meter to the j-th load point.

[0098] The first judgment module is used to judge whether If it is satisfied, the set update module is activated, otherwise the second judgment gate module is activated.

[0099] The set updating module is used to The load error of the smart meter is X j Merge into set Z A , in a ij ∈S, when X j Merge into set Z B , in a ij ∈V, X j Merge into set Z C .

[0100] The second judgment gate module is used to judge whether i>k and j>q are satisfied. If so, the output set Z is A , Z B , Z C , otherwise the first judgment module is activated to perform the next perception cycle, where k=m, q=n+2.

[0101] The risk level determination module is used to determine the risk level based on X j and set Z A , Z B , Z C Determine the risk level of smart meters, including:

[0102] If X j Not in Z A , Z B , Z C If the smart meter is in the union of , it is judged as a low-risk smart meter;

[0103] If X j In the collection Z A If the value is medium, it is judged as a medium-risk smart meter;

[0104] If X j In Z B , Z C If the number of meters is in the same concentration, it is judged as a high-risk smart meter;

[0105] If X j In Z B , Z C If the smart meter is in the intersection of , it is judged as a very high-risk smart meter.

Claims

1. A smart meter error perception evaluation method based on big data, characterized in that: The method comprises: S1. Create three empty sets Z A , Z B , Z C , and based on the first inspection and operation test data of the smart meter, the basic error matrix V = {e ij }, where i = 1, 2, ..., m, j = 1, 2, ..., n, m is the number of smart meters, and n is the number of load points; S2, calculate the average error matrix based on the basic error matrix V Variance matrix S, perception coefficient a ij , the average value of the perception matrix Perception matrix variance S j ', where a ij , S j ' is calculated based on the following formula: In the above formula, V k ' ,q is the perception matrix, a ij is the perception coefficient of the i-th smart meter to the j-th load point; S3. Determine whether If satisfied, go to S4, otherwise go to S5; S4. If Then the load error of the smart meter is j Merge into set Z A If a ij ∈S, then X j Merge into set Z B If a ij ∈V, then X j Merge into set Z C ; S5. Determine whether i>k and j>q. If so, output set Z. A , Z B , Z C , otherwise return to S3 for the next perception cycle, where k = m, q = n + 2; S6, based on X j and set Z A , Z B , Z C Determine the risk level of smart meters.

2. According to the big data-based smart meter error perception evaluation method of claim 1, it is characterized in that: Said S is calculated based on the following formula: In the above formula, is the average error of the i-th smart meter, S i is the variance of the i-th smart meter.

3. The big data-based smart meter error perception and evaluation method according to claim 1 or 2, characterized in that: In S6, the smart meter risk level determination includes: If X j Not in Z A , Z B , Z C If the smart meter is in the union of , it is judged as a low-risk smart meter; If X j In the collection Z A If the value is medium, it is judged as a medium-risk smart meter; If X j In Z B , Z C If the number of meters is in the same concentration, it is judged as a high-risk smart meter; If X j In Z B , Z C If the smart meter is in the intersection of , it is judged as a very high-risk smart meter.

4. A smart meter error perception and evaluation system based on big data, characterized in that: The system includes a set creation module, a perception calculation module, a first judgment module, a set update module, a second judgment gate module, and a risk level determination module; The set creation module is used to create three empty sets Z A , Z B , Z C , and based on the first inspection and operation test data of the smart meter, the basic error matrix V = {e ij }, where i = 1, 2, ..., m, j = 1, 2, ..., n, m is the number of smart meters, and n is the number of load points; The perception calculation module is used to calculate the average error matrix based on the basic error matrix V in turn. Variance matrix S, perception coefficient a ij , the average value of the perception matrix Perception matrix variance S j ′, where a ij , S j ' is calculated based on the following formula: In the above formula, V k ' ,q is the perception matrix, a ij is the perception coefficient of the i-th smart meter to the j-th load point; The first judgment module is used to judge whether If satisfied, the set update module is activated, otherwise the second judgment gate module is activated; The set updating module is used to The load error of the smart meter is X j Merge into set Z A , in a ij ∈S, when X j Merge into set Z B , in a ij ∈V, X j Merge into set Z C ; The second judgment gate module is used to judge whether i>k and j>q are satisfied. If so, the output set Z is A , Z B , Z C , otherwise activate the first judgment module to perform the next perception cycle, where k=m, q=n+2; The risk level determination module is used to determine the risk level based on X j and set Z A , Z B , Z C Determine the risk level of smart meters.

5. The big data-based smart meter error perception and evaluation system according to claim 4 is characterized in that: The perception calculation module is calculated based on the following formula S: In the above formula, is the average error of the i-th smart meter, S i is the variance of the i-th smart meter.

6. The big data-based smart meter error perception and evaluation system according to claim 4 is characterized in that: The risk level determination module determines the risk level of the smart meter based on the following method: If X j Not in Z A , Z B , Z C If the smart meter is in the union of , it is judged as a low-risk smart meter; If X j In the collection Z A If the value is medium, it is judged as a medium-risk smart meter; If X j In Z B , Z C If the number of meters is in the same concentration, it is judged as a high-risk smart meter; If X j In Z B , Z C If the smart meter is in the intersection of , it is judged as a very high-risk smart meter.

Images