A user phase sequence identification method, system and device

Abstract

The application discloses a user phase sequence identification method, system and device in the field of smart grid, and the method comprises the following steps: collecting the injection current of the three-phase bus at the low-voltage side of a distribution transformer at multiple time points and the user load current recorded by a user electric meter; taking the injection current of the three-phase bus as the dependent variable and taking the user load current as the independent variable to establish a multiple linear regression model; performing significance test on the regression coefficient of the user load current in the model, and according to the test result, using a stepwise regression algorithm to screen the user load current of the multiple linear regression model to obtain a significant user load current subset of different phase sequences; introducing a significance correction factor to correct and identify the significant user load current subset to obtain a significant user load current subset of different phase sequences after correction, which is a user set of the same phase sequence respectively. The application reduces the complexity of the distribution network and the input cost, and improves the user phase sequence identification accuracy.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of smart grid, more particularly, to a user phase sequence identification method, system and device. BACKGROUND

[0002] In the low-voltage distribution area of China, as shown in the drawings, the user side is mostly single-phase load, and a single-phase user meter is generally connected to different phase lines (A / B / C) and a neutral line through a downlead, and the connection of different phase lines determines the user phase sequence. Due to the randomness of user power consumption, three-phase imbalance phenomenon is prone to occur in the low-voltage distribution area, and therefore it is necessary to identify the phase sequence of the user to better maintain the three-phase balance of the low-voltage distribution area. Figure 1

[0003] The existing user phase sequence identification method mostly sends and receives characteristic signals of other communication devices through the communication devices installed at the user meter or the line branch point, and distinguishes the connection relationship of the user based on the physical isolation principle of the signal path to identify the user phase sequence.

[0004] However, although the above method can achieve user phase sequence identification, it needs to install a large number of communication devices at the user meter or the line branch point, which will cause the defects of complex distribution network and high investment cost; in addition, the communication data driven method adopted by the above method needs to rely on the size of the regression coefficient value to judge the user phase sequence, and there are measurement error and model error, which will also cause the defect of low identification accuracy. SUMMARY

[0005] In order to overcome the defects of complex distribution network, high investment cost and low identification accuracy in the existing user phase sequence identification method, the present application provides a user phase sequence identification method, system and device.

[0006] To solve the above technical problems, the technical solutions of the present application are as follows:

[0007] In a first aspect, the present application provides a user phase sequence identification method, comprising:

[0008] Collecting the injection current of the three-phase bus at the low-voltage side of the distribution transformer at multiple times, and the user load current recorded by the user meter.

[0009] Taking the injection current of the three-phase bus as the dependent variable and the user load current as the independent variable, a multiple linear regression model is established.

[0010] Performing significance test on the regression coefficient of the user load current in the multiple linear regression model, and according to the test result, using stepwise regression algorithm to screen the user load current of the multiple linear regression model to obtain a significant user load current subset of different phase sequences. ​

[0011] The modified significance correction factor is introduced to correct and identify the significant user load current subset, thereby obtaining the modified significant user load current subset of different phase sequences, i.e., the user set of the same phase sequence.

[0012] In a second aspect, the present application further provides a user phase sequence identification system, comprising:

[0013] The acquisition module is configured to acquire the injection current of the three-phase bus at the low-voltage side of the distribution transformer at multiple time points and the user load current recorded by the user meter.

[0014] The model construction module is configured to establish a multiple linear regression model by taking the injection current of the three-phase bus as the dependent variable and the user load current as the independent variable.

[0015] The inspection module is configured to perform significance inspection on the regression coefficient of the user load current in the multiple linear regression model.

[0016] The screening module is configured to perform stepwise regression screening on the user load current of the multiple linear regression model according to the inspection result of the inspection module, thereby obtaining a significant user load current subset of different phase sequences.

[0017] The correction and identification module is configured to introduce a significance correction factor to correct and identify the significant user load current subset, thereby obtaining the modified significant user load current subset of different phase sequences, i.e., the user set of the same phase sequence.

[0018] In a third aspect, the present application further provides a computing device, comprising a memory and a processor. The memory is configured to store computer executable instructions, and the processor is configured to execute the computer executable instructions to implement the steps of the user phase sequence identification method.

[0019] Compared with the prior art, the technical scheme of the present application has the beneficial effects that: the present application can ensure higher identification accuracy under the influence of various errors by judging the user phase sequence through the significance inspection of the independent variable regression coefficient and screening the significant independent variable by using the stepwise regression algorithm, and the phase sequence identification of the low-voltage distribution area user can be completed only by relying on the meter data without the need for additional installation of other devices, thereby reducing the complexity of the distribution network and the investment cost. BRIEF DESCRIPTION OF DRAWINGS

[0020] Figure 1 It is a schematic diagram of the low-voltage distribution area in the background art.

[0021] Figure 2 It is a flowchart of the user phase sequence identification method.

[0022] Figure 3Intersection diagram for significant subsets of phase-specific independent variables.

[0023] Figure 4 Power grid structure diagram for phase A.

[0024] Figure 5 Architecture diagram of the user phase sequence identification system. DETAILED DESCRIPTION

[0025] The accompanying drawings are only for illustrative purposes and should not be construed as limiting the patent;

[0026] The technical solutions of the present application will be further described below in combination with the drawings and examples.

[0027] Example 1

[0028] Please refer to Figure 2 The embodiment provides a user phase sequence identification method, which comprises:

[0029] Collecting the injection current of the three-phase bus at the low-voltage side of the distribution transformer at multiple time points, and the user load current recorded by the user electric meter.

[0030] Establishing a multiple linear regression model with the injection current of the three-phase bus as the dependent variable and the user load current as the independent variable.

[0031] Performing a significance test on the regression coefficient of the user load current in the multiple linear regression model, and using a stepwise regression algorithm to screen the user load current of the multiple linear regression model according to the test result, to obtain a significant user load current subset of different phase sequences.

[0032] Introducing a significant correction factor to correct and identify the significant user load current subset, to obtain a significant user load current subset of different phase sequences after correction, i.e., a user set of the same phase sequence.

[0033] In the specific implementation process, the user phase sequence is determined by the significance test of the regression coefficient of the independent variable, and the significant independent variables are screened using the stepwise regression algorithm, which can ensure higher identification accuracy under various error influences, and does not require additional installation of other equipment. Only by relying on the electric meter data, the phase sequence identification of the low-voltage distribution area users can be completed, reducing the complexity and investment cost of the power distribution network.

[0034] Example 2

[0035] The embodiment provides a user phase sequence identification method, which comprises:

[0036] Collecting the injection current of the three-phase bus at the low-voltage side of the distribution transformer at multiple time points, and the user load current recorded by the user electric meter.

[0037] In this embodiment, after collecting the three-phase bus injection currents at the low-voltage side of the distribution transformer at multiple time points and the user load currents recorded by each user electric meter, a three-phase bus injection current vector I A ∈ T , B ∈ T , C ∈ T and a user load current matrix I M ∈ T×N are constructed using the three-phase bus injection currents and the user load currents, respectively. The three-phase bus injection current vector and the user load current matrix are used to construct a multiple linear regression model, and the specific expression is as follows:

[0038]

[0039]

[0040]

[0041]

[0042] wherein I A represents the A-phase bus injection current vector, I Ai represents the A-phase bus injection current at the i-th time point, I B represents the B-phase bus injection current vector, I Bi represents the B-phase bus injection current at the i-th time point, I C represents the C-phase bus injection current vector, I Ci represents the C-phase bus injection current at the i-th time point, H = {1, 2, …, T} is a time set, T represents the total number of recorded time points, L = {1, 2, …, N} is a user set, N represents the total number of users, I M represents the user load current matrix, I Mij represents the load current of the j-th user at the i-th time point.

[0043] A multiple linear regression model is established with the three-phase bus injection currents as dependent variables and the user load currents as independent variables.

[0044] In this embodiment, a multiple linear regression model is established with each phase bus injection current as the dependent variable Y and the user load current as the independent variable X based on the observation data, and the expression is as follows:

[0045] Y = Xβ + e

[0046] Y = [I A , I B , I C] T

[0047]

[0048] Wherein, Y is dependent variable three-phase bus injection current, X is independent variable user load current, β is independent variable regression coefficient vector, and e is error vector.

[0049] In the embodiment, the regression coefficients of the user load current in the multiple linear regression model are subjected to significance test, and according to the test result, the user load current of the multiple linear regression model is filtered by using a stepwise regression algorithm to obtain a significant user load current subset of different phase sequences.

[0050] According to the principle of energy conservation, the load of each phase bus should theoretically equal the sum of the user load of the same phase sequence. If the load of each phase bus is taken as the dependent variable and the user load is taken as the independent variable, it can be considered from the statistical point of view that the load of a certain phase bus linearly depends on the user load of the same phase sequence, that is, the regression coefficient of the independent variable corresponding to all users of the phase should be significantly not 0. From the perspective of hypothesis testing, it is equivalent to testing whether the null hypothesis H0 is accepted.

[0051]

[0052] In the embodiment, the regression coefficients of the independent variables in the multiple linear regression model are subjected to significance test by using a variance homogeneity test, and the obtained P0 is called the significance test P value of the regression coefficient β = 0 of the independent variable. The smaller the P0 value is, the smaller the probability of β = 0 is, that is, the stronger the significance of the dependent variable is, and vice versa.

[0053] In the embodiment, according to the test result, the user load current of the multiple linear regression model is filtered by using a stepwise regression algorithm to obtain a significant user load current subset of different phase sequences, which specifically includes:

[0054] Step A: respectively setting a significant threshold λ entry of the introduced user load current and a significant threshold λ remove of the eliminated user load current, and λ entry < λ remove .

[0055] Step B: using a variance homogeneity test to subject the regression coefficient of any user load current in the multiple linear regression model to significance test, and calculating the significance test P value P0 of the regression coefficient of the user load current being 0.

[0056] Step C: when P0 < λ entry , the user load current is introduced into the multiple linear regression model as one of the independent variables. When P0 > λ removeWhen this happens, the user's load current will be excluded.

[0057] Step D: Repeat steps B-C until the significance test is completed for all user load currents in the multiple linear regression model. Construct a subset of significant user load currents with different phase sequences using the user load currents introduced into the multiple linear regression model.

[0058] Stepwise regression is an algorithm that iteratively finds the subset of independent variables that best explains the observed values ​​of the dependent variable, based on the results of significance tests of the independent variables. The main approach of stepwise regression is to introduce variables one by one, and then test whether the inclusion condition (P0 < λ) is met. entry If a new variable is introduced, then the new variable is introduced. For each new variable introduced, the existing variables in the equation are tested one by one. Those variables deemed to satisfy the insignificance elimination condition (P0>λ) are removed. remove If the old variable is not significant, it is removed to ensure that every variable in the resulting subset of independent variables is significant. This process is repeated several times until no new variables can be introduced. To avoid falling into an infinite loop of "introducing-removing-introducing" the same variable, a significance threshold λ for introducing independent variables is generally required. entry It must be less than the significance threshold λ for removing independent variables. remove , that is, λ entry <λ remove .

[0059] A significance correction factor is introduced to correct the subset of significant independent variables, and the user phase sequence is identified based on the corrected subset of significant independent variables.

[0060] In this embodiment, the subset X of significant user load currents in phase A can be obtained by using a stepwise regression algorithm. inA The significant user load current subset X of phase B inB and the C-phase significant user load current subset X inC Its specific expression is as follows:

[0061] X inA ={x A(1) ,x A(2) ,…,x A(i) ,,,,,x A( n A)}n A ∈L

[0062] X inB ={x B(1) ,x B(2) ,…,x B(i) ,,,,,x B( n B)}n B ∈L

[0063] X inC ={x C(1) ,x C(2) ,…,x C(i) ,,,,,x C( n C}n C ∈L

[0064] Where, x A(i) x B(i) and x C(i) These represent the i-th user whose phase sequence is identified as phases A, B, and C, respectively, and n. A n B and n C Let L represent the number of users included in the significant user load current subsets corresponding to phases A, B, and C, respectively. Let L = {1, 2, ..., N} be the user set, and N represent the total number of users.

[0065] Considering the impact of various errors and significance threshold settings, X inA X inB and X inC There may be overlap between them, such as Figure 3 As shown, Figure 3 The diagram shows the intersection of the significant subsets of independent variables for each phase. However, according to physical constraints, it is impossible for a single-phase user to be connected to different phase lines at the same time. Therefore, the results obtained from stepwise regression need to be corrected.

[0066] According to the significance test theorem, for users with smaller errors and more obvious load characteristics, the expected value of their independent variable regression coefficient should be closer to 1 and the variance should be smaller, that is, the probability of the coefficient being 1 is greater and the probability of it being 0 is smaller.

[0067] In this embodiment, a significance correction factor is introduced. The introduction of the significance correction factor, which corrects and identifies the subset of significant independent variables, specifically includes:

[0068] The significance correction factor ζ is introduced, and its calculation formula is as follows:

[0069] ζ = ln(P1 / P0)

[0070] Where P1 represents the significance test P-value P1 for the regression coefficient of the independent variable being 1. The smaller the value, the smaller the probability of β = 1. Its calculation method can be referred to P0.

[0071] By definition, the numerical range of ζ is (-∞, +∞). When P1 > P0, ζ > 0. In the extreme case, when P1 = 1 and P0 = 0, ζ = +∞ and β = 1; conversely, when P1 > P0, ζ > 0. <P jWhen ζ < 0, in the extreme case, when P1 = 0 and P0 = 1, ζ = -∞ and β = 0; when P1 = P0, ζ = 0 and β = 0 or 1 have the same probability, and the uncertainty is the highest.

[0072] For a significant subset of independent variables X inA X inB or X inC If the user set of phases A, B, or C has a correction factor ζ for a certain user j. V(j) If ≤0, V=A,B,C, then user j is removed from the significant independent variable subset X. inA X inB or X inC Remove from the middle;

[0073] For a significant subset of independent variables X inA and X inB The intersection X inAB If the correction factor ζ of a certain user j in the user set of phase A is... A(j) And in the user set of phase B, there exists a correction factor ζ for a certain user j. B(j) , so that ζ A(j) >ζ B(j) If the value is greater than 0, then user j will be removed from the subset X of the significant independent variables. inB Remove from the middle; if the correction factor ζ of a certain user j in the user set of phase A is... A(j) And in the user set of phase B, there exists a correction factor ζ for a certain user j. B(j) , so that ζ B(j) >ζ A(j) If the value is greater than 0, then user j will be removed from the subset X of the significant independent variables. inA Remove from the middle;

[0074] For a significant subset of independent variables X inA and X inC The intersection X inAC If the correction factor ζ of a certain user j in the user set of phase A is... A(j) And in the user set of phase C, there exists a correction factor ζ for a certain user j. C(j) , so that ζ A(j) >ζ C(j) If the value is greater than 0, then user j will be removed from the subset X of the significant independent variables. inC Remove from the middle; if the correction factor ζ of a certain user j in the user set of phase A is... A(j) And in the user set of phase C, there exists a correction factor ζ for a certain user j. C(j) , so that ζ C(j) >ζ A(j) If the value is greater than 0, then user j will be removed from the subset X of the significant independent variables. inA Remove from the middle;

[0075] For a significant subset of independent variables X inB and X inC The intersection X inBC If the correction factor ζ of a certain user j in the user set of phase B is... A(j) And in the user set of phase C, there exists a correction factor ζ for a certain user j. C(j) , so that ζ B(j) >ζ C(j) If the value is greater than 0, then user j will be removed from the subset X of the significant independent variables. inC Remove from the middle; if the correction factor ζ of a certain user j in the user set of phase B is... B(j) And in the user set of phase C, there exists a correction factor ζ for a certain user j. C(j) , so that ζ C(j) >ζ B(j) If the value is greater than 0, then user j will be removed from the subset X of the significant independent variables. inB Remove from the list.

[0076] In this embodiment, after identifying the user phase sequence based on the modified subset of significant independent variables, for users whose phase sequence could not be identified due to weak significance, voltage correlation analysis or on-site investigation is used to identify the phase sequence.

[0077] In one specific embodiment, a simulation test system with a total of 63 users is established, using a typical low-voltage distribution substation network structure in China as a reference. The users in phase A are numbered 1-21, those in phase B are numbered 22-42, and those in phase C are numbered 43-63. For brevity, as follows... Figure 4 As shown, this is the power supply network structure diagram for phase A. The main line type is BLV-150, and the down conductor type is BLV-50. The lengths of each line are as follows: Figure 4 As shown.

[0078] Taking actual user load data collected in a certain region of Guangdong, China as an example, the collection time interval is 15 minutes, the time period is 2 days, and there are a total of T = 192 time sections. Using the low-voltage side of the distribution transformer as the balancing nodes, multi-section power flow simulation calculations are performed to obtain the injected current data for each phase bus. If a significance threshold λ is set... remove =2λ entry =0.05, with the bus current of each phase as the dependent variable and the user load current as the independent variable, the preliminary user phase sequence identification results are shown in Table 1 before the introduction of correction factors. The user serial number with an underline indicates the user with the incorrect phase sequence identification.

[0079] Table 1 Preliminary User Phase Sequence Identification Results (X) inA X inB X inC )

[0080]

[0081]

[0082] Further, in order to improve the recognition accuracy and avoid violation of physical constraints, a significance correction factor ζ of each phase independent variable in the set is calculated. The value of the significance correction factor of each phase independent variable in the set is shown in Table 2, wherein the user number with an underline indicates a user with phase sequence recognition error.

[0083] Table 2 Significance correction factor of each phase independent variable in the set

[0084]

[0085]

[0086] After the results in Table 1 are corrected based on the significance correction factor ζ, the results shown in Table 3 are obtained, wherein the user number with an underline is a user with phase sequence recognition error.

[0087] Table 3 Corrected phase sequence recognition results (X’ inA , X’ inB , X’ inC )

[0088]

[0089] As can be seen from Table 3, the corrected results reduce the misrecognition and make a second judgment on the phase sequence of the multi-phase cross users, greatly improving the precision rate.

[0090] For users (13, 17, 20, 23, 25, 32) whose phase sequence cannot be recognized due to weak significance, on-site investigation or voltage correlation analysis can be used for phase sequence recognition.

[0091] Embodiment 3

[0092] Please refer to Figure 5 , the embodiment proposes a user phase sequence recognition system, which comprises a collection module, a model construction module, a verification module, a screening module and a correction recognition module.

[0093] In the specific implementation process, the acquisition module acquires the injection currents of the three-phase busbar at the low-voltage side of the distribution transformer at multiple moments, and the user load current recorded by the user meter; the model construction module establishes a multiple linear regression model with the injection currents of the three-phase busbar as dependent variables and the user load current as an independent variable; the test module performs significance test on the regression coefficients of the user load current in the multiple linear regression model; the screening module performs stepwise regression screening on the user load current of the multiple linear regression model according to the test result of the test module, and obtains a significant user load current subset of different phase sequences; and the correction and identification module introduces a significance correction factor to correct and identify the significant user load current subset, and obtains a corrected significant user load current subset of different phase sequences, i.e., a user set of the same phase sequence.

[0094] The phase sequence of the user is determined by the significance test on the regression coefficients of the independent variables, and the stepwise regression algorithm is used to screen the significant independent variables, so that higher identification accuracy can be ensured under various error influences, and the phase sequence identification of the low-voltage distribution area user can be completed only by relying on the meter data without the need of installing other devices, thereby reducing the complexity and investment cost of the power distribution network.

[0095] The terms describing the positional relationship in the drawings are only used for illustrative description, and should not be understood as a limitation to the patent;

[0096] Obviously, the above embodiments of the present application are only examples for clearly illustrating the present application, and are not intended to limit the implementation manners of the present application. Based on the above description, other different forms of changes or variations can be made by those skilled in the art. Here, all the implementation manners are not required or can not be exhausted. Any modification, equivalent replacement and improvement made within the spirit and principle of the present application should be included in the protection scope of the claims of the present application.

Claims

1. A user phase sequence identification method, characterized by, The method comprises the following steps: Collecting the injection currents of the three-phase bus at the low-voltage side of the distribution transformer at multiple time points, and the user load currents recorded by the user meter; Establishing a multiple linear regression model by taking the injection currents of the three-phase bus as dependent variables and the user load currents as independent variables; Performing significance test on the regression coefficients of the user load currents in the multiple linear regression model, and screening the user load currents in the multiple linear regression model according to the test results by using a stepwise regression algorithm to obtain significant user load current subsets of different phase sequences; Introducing a significance correction factor to correct and identify the significant user load current subsets to obtain corrected significant user load current subsets of different phase sequences, i.e., user sets of the same phase sequence; Wherein, the significance correction factor ζ is calculated according to the following formula: ζ = ln (P1 / P0) In the formula, P1 represents the significance test P value of the user load current regression coefficient being 1, and P0 represents the significance test P value of the user load current regression coefficient being 0.

2. The user phase identification method of claim 1, wherein, After the three-phase bus injection currents and the user load currents are collected, the method further comprises: constructing a three-phase bus injection current vector by using the three-phase bus injection currents respectively, and constructing a user load current matrix by using the user load currents, wherein the three-phase bus injection current vector and the user load current matrix are used to construct a multiple linear regression model, and the specific expression is as follows: wherein I A denotes the A-phase bus injection current vector, I Ai denotes the A-phase bus injection current at the i-th time instant, I B denotes the B-phase bus injection current vector, I Bi denotes the B-phase bus injection current at the i-th time instant, I C denotes the C-phase bus injection current vector, I Ci denotes the C-phase bus injection current at the i-th time instant, H = {1, 2,..., T} is the time set, T denotes the total number of recorded time instants, L = {1, 2,..., N} is the user set, N denotes the total number of users, I M denotes the user load current matrix, I Mij denotes the load current of the j-th user at the i-th time instant.

3. The user phase identification method of claim 2, wherein, The expression of the multiple linear regression model is as follows: Y = Xβ + e Y = [I A ,I B ,I C ] T Wherein, Y is the dependent variable three-phase bus injection current, X is the independent variable user load current, β is the regression coefficient vector of the independent variable, and e is the error vector.

4. The user phase identification method of claim 1, wherein, The method further comprises the following steps: Step A: Set the significance threshold λ entry and the significance threshold λ remove for rejecting the user load current, respectively, and λ entry < λ remove ; Step B: performing significance test on the regression coefficient of any user load current in the multiple linear regression model by using a variance homogeneity test, and calculating the significance test P value P0 of the user load current regression coefficient being 0; Step C: When P0< λ entry introduce the user load current into the multiple linear regression model as one of the independent variables; when P0> λ remove eliminate the user load current; Step D: repeating steps B-C until the significance test of all user load currents in the multiple linear regression model is completed, and constructing significant user load current subsets of different phase sequences by using the user load currents introduced into the multiple linear regression model.

5. The user phase identification method of claim 1, wherein, The significant user load current subsets of the different phases include an A-phase significant user load current subset X inA , a B-phase significant user load current subset X inB , and a C-phase significant user load current subset X inC , which are specifically expressed as follows: X inA = {x A(1) , x A(2) ,..., x A(i) ,..., x A( n A)}n A ∈ L X inB = {x B(1) , x B(2) ,..., x B(i) ,..., x B( n B)}n B ∈ L X inC = {x C(1) , x C(2) ,..., x C(i) ,..., x C( n C}n C ∈ L where x A(i) , x B(i) and x C(i) represent the ith user whose phase sequence is identified as phase A, B and C, respectively, n A , n B and n C are the number of users included in the sub-set of significant user load current corresponding to phase A, B and C, respectively, L = {1, 2, …, N} is the user set, and N represents the total number of users.

6. The user phase identification method of claim 5, wherein, The method further comprises the following steps: For the subset of significant user load currents X inA , X inB , or X inC , if the set of users of phase A, B or C has a correction factor ζ V(j) ≤ 0 for some user j, V = A, B, C, then user j is removed from the subset of significant user load currents X inA , X inB , or X inC ; For the significant user load current subset X inA and X inB The intersection X inAB If the correction factor ζ of a certain user j in the user set of phase A is... A(j) And in the user set of phase B, there exists a correction factor ζ for a certain user j. B(j) , so that ζ A(j) >ζ B(j) If the value is greater than 0, then user j will be removed from the significant user load current subset X. inB Remove from the middle; if the correction factor ζ of a certain user j in the user set of phase A is... A(j) And in the user set of phase B, there exists a correction factor ζ for a certain user j. B(j) , so that ζ B(j) >ζ A(j) If the value is greater than 0, then user j will be removed from the significant user load current subset X. inA Remove from the middle; For the significant user load current subset X inA and X inC The intersection X inAC If the correction factor ζ of a certain user j in the user set of phase A is... A(j) And in the user set of phase C, there exists a correction factor ζ for a certain user j. C(j) , so that ζ A(j) >ζ C(j) If the value is greater than 0, then user j will be removed from the significant user load current subset X. inC Remove from the middle; if the correction factor ζ of a certain user j in the user set of phase A is... A(j) And in the user set of phase C, there exists a correction factor ζ for a certain user j. C(j) , so that ζ C(j) >ζ A(j) If the value is greater than 0, then user j will be removed from the significant user load current subset X. inA Remove from the middle; For the significant user load current subset X inB and X inC The intersection X inBC If the correction factor ζ of a certain user j in the user set of phase B is... A(j) And in the user set of phase C, there exists a correction factor ζ for a certain user j. C(j) , so that ζ B(j) >ζ C(j) If the value is greater than 0, then user j will be removed from the significant user load current subset X. inC Remove from the middle; if the correction factor ζ of a certain user j in the user set of phase B is... B(j) And in the user set of phase C, there exists a correction factor ζ for a certain user j. C(j) , so that ζ C(j) >ζ B(j) If the value is greater than 0, then user j will be removed from the significant user load current subset X. inB Remove from the list.

7. The user phase identification method according to any one of claims 1 to 6, characterized in that, The method further comprises the following steps:

8. The user phase identification method according to any one of claims 1 to 6, characterized in that, The method further comprises the following steps:

9. A user phase sequence identification method system characterized by, ​ The collection module is configured to collect injection currents of a three-phase bus at a low-voltage side of a distribution transformer at multiple time points and user load currents recorded by user electricity meters; The model construction module is configured to establish a multiple linear regression model by taking the injection currents of the three-phase bus as dependent variables and the user load currents as independent variables; The inspection module is configured to perform significance inspection on regression coefficients of the user load currents in the multiple linear regression model; The screening module is configured to perform stepwise regression screening on the user load currents in the multiple linear regression model according to an inspection result of the inspection module, to obtain a subset of significant user load currents of different phase sequences; The correction and identification module is configured to introduce a significant correction factor to perform correction and identification on the subset of significant user load currents, to obtain a subset of significant user load currents of different phase sequences after correction, i.e., user sets of the same phase sequence. The significant correction factor ζ is calculated according to the following formula: ζ = ln(P1 / P0) In the formula, P1 represents a significance test P value of a user load current regression coefficient of 1, and P0 represents a significance test P value of a user load current regression coefficient of 0.

10. A computing device comprising: A memory and a processor; The memory is configured to store computer executable instructions, and the processor is configured to execute the computer executable instructions to implement steps of the user phase sequence identification method in any one of claims 1 to 8.

Images