Wiring error electric leakage user detection method and device

Abstract

The invention provides a wiring error electric leakage user detection method and device. The detection method comprises the following steps: (A) in an acquisition time period, acquiring residual current of a detected transformer area and load current of each user in the detected transformer area at a set sampling interval; constructing a residual function, wherein the residual function is a residual sum of squares of a sampling value and an estimated value of the residual current of the tested transformer area in the acquisition time period; (B) estimating a complex weight coefficient of the load current of each user in the multiple linear regression equation about the residual current of the tested transformer area by minimizing the value of the residual function; and (C) if the amplitude of the estimated complex weight coefficient of the jth user is greater than or equal to a preset amplitude threshold, judging that the jth user is a wiring error electric leakage user, otherwise, judging that the jth user is a normal user. According to the method, the size of the complex weight coefficient of the load current of each user about the residual current of the transformer area is iteratively calculated in the feasible region, and the obtained identification result is more accurate and reliable.

Description

Technical Field

[0001] The present invention relates to the field of power system fault analysis, in particular to the identification and location of leakage faults caused by incorrect wiring of user distribution boxes in low-voltage distribution networks, and particularly to a method for detecting users with incorrect wiring and leakage based on the primal-dual interior point method. Background Art

[0002] Low-voltage distribution areas are the basic units of the distribution network, serving many users and having complex feeders. During continuous new construction, renovation, maintenance, and demolition processes, the distribution lines are a mixture of new and old, and the operating environment is harsh. This has caused great difficulties to the daily inspection work of low-voltage distribution operation and maintenance personnel. Moreover, inexperienced and less professional electricians often operate improperly in the face of complex distribution environments, damaging the normal operation of the lines and creating new safety hazards by damaging the original lines. Therefore, in engineering, in addition to leakage due to damaged line insulation, incorrect wiring of the neutral and ground wires in indoor distribution boxes is also an important reason for abnormal residual current in the area and the difficulty in putting into use the residual current protection. When such a fault occurs, if the user's household protection fails or is not put into operation, the leakage fault will exist for a long time, affecting the residual current in the area while not affecting the normal power consumption of the user. And this type of fault is very hidden and does not damage the insulation characteristics between the live wire and the ground wire or between the neutral wire and the ground wire like an insulation leakage fault. Therefore, it is even more difficult to detect. When this leakage fault occurs in a user with normal household protection, due to the difficulty in identifying and detecting the fault, in order to ensure power consumption, the leakage protection device will be withdrawn from use without authorization, posing a great potential safety hazard for power consumption. On the other hand, when a user has incorrect wiring and leakage, the load current returns to the transformer neutral point through the ground wire, causing the ground wire to carry a large load current for a long time, accelerating the oxidation of the grounding resistance and increasing the risk of electric shock. Its fault voltage may also bypass other users' household protection through the ground wire and pose a threat to other users. To ensure the safety of power consumption in the low-voltage distribution network, it is urgent to identify and detect such faults.

[0003] During the construction of a low-voltage transparent distribution network with the large-scale promotion of HPLC smart meters, the use of a residual current monitoring module configured in the intelligent distribution transformer terminal can monitor the residual current in the area, providing rich data support for the analysis of the characteristics of multiple electrical quantities in the area before and after the fault.

[0004] However, in the prior art, only the correlation between the residual current in the area and the user load current in the real number domain is considered, and it is impossible to effectively identify users with incorrect wiring and leakage. Summary of the Invention

[0005] The problem to be solved by the present invention is to provide a more accurate and reliable method for detecting users with incorrect wiring and leakage in view of the problem that the prior art only considers the correlation between the residual current in the area and the user load current in the real number domain and thus cannot effectively identify users with incorrect wiring and leakage.

[0006] To solve the above technical problems, the technical solution adopted by the present invention is: A method for detecting users with wiring errors and leakage, including:

[0007] (A) During the acquisition time period, the residual current of the measured substation area and the load currents of each user in the measured substation area are acquired at a set sampling interval; a residual function is constructed, and the residual function is the sum of the squared residuals of the sampled values and the estimated values of the residual current of the measured substation area during the acquisition time period;

[0008] (B) By making the value of the residual function minimum, the complex weight coefficients of each user's load current with respect to the residual current of the measured substation area in the multiple linear regression equation are estimated;

[0009] (C) If the magnitude of the complex weight coefficient of the j-th user obtained by estimation is greater than or equal to the preset magnitude threshold, it is determined that the j-th user is a user with wiring errors and leakage; otherwise, it is determined that the j-th user is a normal user;

[0010] where j ∈ [1, N], and N represents the number of users in the measured substation area.

[0011] In the present invention, according to electrical prior knowledge, the correlation between the residual current of the substation area and the load currents of each user is considered. By setting a residual function, more accurate and reliable values of the complex weight coefficients of the user load currents are calculated within the feasible region, providing more accurate and reliable guiding information for identifying and locating users with wiring error and leakage faults.

[0012] In the above technical solution, the constructed residual function is:

[0013]

[0014] where, is the complex weight coefficient of the load current of the j-th user with respect to the residual current of the measured substation area; is the i-th sampled value of the residual current of the measured substation area, is the i-th sampled value of the load current of the j-th user; i ∈ [1, M], and M represents the number of sampled values of the residual current of the measured substation area; M > N.

[0015] In the above technical solution, in step (B), the specific method for estimating the complex weight coefficients of each user's load current with respect to the residual current of the measured substation area by making the value of the residual function minimum is:

[0016] Solve the constructed second-order cone programming problem to obtain the value of the complex weight coefficient of the load current of the j-th user in the measured substation area with respect to the residual current of the measured substation area;

[0017] The constructed second-order cone programming problem is: subject to Under the condition of, minimize the value of E'(β);

[0018] Among them, Residual current matrix Denote the conjugate transpose matrix of I r , β H Denote the conjugate transpose matrix of β, load current matrix

[0019] In the above technical solution, the primal-dual interior point method is used to solve the constructed second-order cone programming problem.

[0020] In the above technical solution, the specific method for using the primal-dual interior point method to solve the constructed second-order cone programming problem is: (B1) Perform iterative operations using the following formula:

[0021]

[0022] Among them:

[0023]

[0024] J(β k , λ k ) represents the Jacobian matrix of F(β k , λ k ); Denote the partial derivative of in ;

[0025]

[0026]

[0027]

[0028]

[0029] Among them, is the constructed Lagrangian function, (β k ) H represents the conjugate transpose matrix of β k , Denote the amplitude of, Denote the conjugate transpose matrix of I L ; α_p is the primal step size, α_d is the dual step size, β k is the k-th complex weight coefficient update matrix; λ k is the k-th dual variable matrix; Δβ k is the k-th first increment matrix, Δλk is the k-th second increment matrix; k = 0, 1, ……; β 0 each element in, λ 0 each element in, μ 0 , σ are all preset values, μ 0 , σ have a value range of (0, 1], β 0 each element in is equal and has a value range of (0, 1], λ 0 each element in is equal and has a value range of (0, 1];

[0030] (B2) Determine whether the iteration stop condition is satisfied. If the determination result is yes, stop the iteration and use β k+1 as the obtained complex weight coefficient β, otherwise, continue the iteration.

[0031] In the above technical solution:

[0032]

[0033] where q = 1, 2, 3; F q1 , F q2 , ……, F qN respectively represent the element in the first column of the q-th row, the element in the second column of the q-th row, ……, the element in the N-th column of the matrix F(β k , λ k ).

[0034] In the above technical solution, the iteration stop condition is:

[0035] The number of iterations reaches the preset number of iteration values; or

[0036] The norm of the residual r(β, λ) k+1 is less than or equal to the preset norm threshold;

[0037] where the residual r(β, λ) k+1 =(r_c k+1 , r_b k+1 , r_s k+1 );

[0038]

[0039] In the above technical solution, the value range of the preset norm threshold is [10 -4 , 10 -3 .

[0040] In the above technical solution, the value range of the preset amplitude threshold is [0.5, 0.9].

[0041] Based on the same inventive concept, the present invention also provides a wiring error and leakage user detection device, which includes a computer device; the computer device is configured or programmed to execute the steps of the above-mentioned wiring error and leakage user detection method. Description of the Drawings

[0042] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0043] Figure 1 It is a flowchart of wiring error and leakage fault identification based on the primal-dual interior point method in the embodiments of the present invention. Detailed Embodiments

[0044] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the drawings of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, rather than all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0045] In this embodiment, the residual current in the substation area refers to the residual current on the distribution transformer side of the substation area. During the analysis period and within the acquisition time period, it is considered that the complex weight coefficient of each user's load current with respect to the residual current in the substation area and the line leakage current are set as constant values.

[0046] In this embodiment, the complex weight coefficient of each user's load current with respect to the residual current in the substation area is iteratively calculated within the feasible region, and the obtained identification result is more accurate and reliable. In this embodiment, considering that the weight coefficient values and changes of each explanatory variable should follow corresponding range limitations according to the physical characteristics of the object, constraints are added to the linear relationship between the residual current in the substation area and each user's load current constructed according to electrical prior knowledge, and the optimal solution that conforms to engineering practice is solved within the feasible region, further improving the accuracy and reliability of the calculation results.

[0047] As Figure 1 shown, for the wiring error and leakage user detection method based on the primal-dual interior point method in this embodiment, the wiring error and leakage user detection steps are as follows:

[0048] A. Obtain the phasor data of the residual current on the distribution transformer side of the substation area and each user's load current at a certain sampling interval within a certain time period, and construct a multiple linear regression equation according to Kirchhoff's current and phasor superposition theorem.

[0049] Limit the value range of the complex weight coefficient according to the mathematical relationship between the user load current and the user leakage current, and set constraint conditions for the obtained multiple linear regression equation.

[0050] Calculate the residual function of the multiple linear regression equation set according to the least square theory, and transform it into a linear optimization problem of a real-valued complex variable function with constraints.

[0051] B. Use the primal-dual interior point method to solve the complex weight coefficient vector that minimizes the residual function value within the feasible region.

[0052] C. Distinguish the normal users and the abnormal users with wiring errors inside the transformer area according to the magnitudes of the obtained complex weight coefficients of each user.

[0053] The sampling interval in step A is second-level sampling, and the number of sample points needs to be greater than the number of users in the transformer area.

[0054] This is because, in essence, this method requires the optimal solution of the residual function of the constructed multiple linear regression equation. Therefore, it is necessary to ensure that the number of sample points is more than the number of users, that is, the number of independent variables participating in the multiple linear regression equation, to ensure that a unique solution can be obtained. Otherwise, it may be only one of the feasible solutions and not necessarily the optimal solution, which may lead to a local optimum. The obtained results will be even more inconsistent with the physical laws. Therefore, it is necessary to ensure that the number of sample points is a little more, more than the number of users.

[0055] First, measure the multicollinearity between the load current data of each user in the transformer area through the magnitude of the variance inflation factor. The number of sample points should meet the prerequisite that the variance inflation factor of each user's load current compared to the load currents of other users is less than 10. This is because, by meeting the above conditions, the obtained results are reliable and the optimal solution can be obtained.

[0056] This step also takes into account the calculation of the optimal solution. Because there is similarity between the user load currents, that is, they are relatively correlated, which is reflected in multicollinearity mathematically. That is, independent variable A can be obtained through a certain transformation of independent variable B. Therefore, there is a mathematical processing method, which is to calculate the variance inflation factor to measure the multicollinearity between independent variables. Only when the variance inflation factor of each independent variable relative to other independent variables is less than 10 can these independent variables be considered to meet the condition of linear independence and there is no multicollinearity, and then the optimal solution can be further obtained. This step is not mentioned in the flowchart and belongs to the data processing and verification part.

[0057] In the multiple linear regression equation in step A, the residual current on the distribution transformer side of the transformer area is the explained variable, the load currents of each user are the explanatory variables, and the leakage current of the power supply line is the constant term. Its formula is:

[0058]

[0059] Among them, is the residual current phasor value, is the phasor value of each user's load current, represents the magnitude of the complex weight coefficient of each user's load current with respect to the residual current. N is the number of users supplied by the transformer substation area, is the phasor value of the leakage current of the power supply line.

[0060] In step A, the constraint conditions for setting the complex weight coefficient of each user's load current with respect to the residual current in the transformer substation area are:

[0061] S.t. |β j | 2 ≤ 1, j ∈ [0, N];

[0062] In step A, the least squares theory is used to calculate the sum of squared residuals function of the actual value and the estimated value of the residual current, and minimize it to estimate the phasor value of the complex weight coefficient of each user's load current in the multiple linear regression equation. The objective function is:

[0063]

[0064] Among them, M is the total number of sample points, is the vector of each user's load current at the i-th moment, β is the vector of the complex weight coefficient of each user's load current, is the phasor value of the residual current in the transformer substation area at the i-th moment.

[0065] This objective function represents a real-valued complex variable function that maps from the complex domain to the real domain, and can be further rewritten as a second-order cone programming problem P(β):

[0066]

[0067] Among them, I r ∈ C M×1 , represents M sampling values of the residual current in the transformer substation area, I L ∈ C M×N , represents the sampling values of N user load currents at M moments.

[0068] The primal-dual interior point method is used to iteratively solve the complex weight coefficient solution that makes problem P optimal within the feasible region.

[0069] During the analysis period, the weight coefficient of each user's load current with respect to the residual current and the line leakage current are set to constant values.

[0070] In step B, the original step size and the dual step size are calculated respectively to update the original variable and the dual variable.

[0071] The optimal complex weight coefficient vector in step B is calculated by the SDP3 solver of the CVX function package in matlab.

[0072] SDP3 is merely an iterative solver for solving the optimal programming problem and is a way to calculate and obtain results. The innovation of this solution lies in adding constraint conditions that conform to electrical prior knowledge to the solution of the optimal coefficients in the multiple regression equation. Because the results obtained by mathematical solutions conform to mathematical laws, but they may not necessarily apply to physical law phenomena, so constraint conditions need to be added. In this way, the analytical solution solving method in the conventional multiple linear regression calculation process is not applicable. Therefore, an iterative solving method combining the primal-dual interior point method is adopted, and the constraint conditions are added, and the two are combined to obtain a more accurate and effective evaluation result. To improve the recognition accuracy of faults, SDP3 is merely an implementation tool.

[0073] The present invention discloses a method for identifying users with leakage faults due to incorrect wiring of the neutral line and ground wire in a low-voltage distribution substation area based on the primal-dual interior point method. Taking the residual current phasor of the substation area as the explained variable and the load current of the users supplied by the substation area as the explanatory variable, a complex multiple linear regression equation set is constructed. According to electrical prior knowledge, constraint conditions for the complex weight coefficients of each user's load current with respect to the residual current of the substation area are set. By the least squares theory, the multiple linear regression problem is transformed into a convex function linear optimization problem, and the primal-dual interior point method is used to calculate the complex weight coefficients of each user's load current with respect to the abnormal residual current of the substation area within the feasible region to distinguish normal users and abnormal users. The present invention further considers the contribution of the user load current to the residual current of the substation area during the wiring error leakage fault from the perspective of engineering practice, effectively improving the accuracy and reliability of identifying users with wiring error faults, providing targeted guidance for indoor leakage detection of users, and ensuring the safety of user electricity consumption.

[0074] According to electrical prior knowledge, the present invention considers the correlation between the residual current of the substation area and the load current of each user, and sets corresponding constraint conditions for the complex weight coefficients in the constructed multiple linear regression equation, so as to calculate more accurate and reliable complex weight coefficient values of the user load current within the feasible region, providing more accurate and reliable guiding information for identifying and positioning users with wiring error leakage faults. In fact, the contribution of the normal user load current to the abnormal residual current of the substation area is very weak. Therefore, both the real part component and the imaginary part component of the complex weight coefficient are small, and its modulus value is small. While all the abnormal user load current is converted into residual current, in fact, the conversion ratio should be 100%. However, due to the existence of errors, the calculation results of the complex weight coefficients of normal users will have a certain range, but not very large, basically below 0.3, while the amplitude of the complex weight coefficients of abnormal users is basically above 0.95. By using 0.9 as the threshold for division, abnormal users and normal users can be distinguished.

[0075] To achieve the above object, the present invention proposes a method for detecting users with wiring error leakage faults based on the primal-dual interior point method.

[0076] For a distribution area with N users supplied by a feeder, when it operates normally, the residual current obtained at the low-voltage side of its distribution transformer comes from the leakage current of the distribution line to the ground and the leakage currents of each user. There is a phasor superposition relationship between each current component and the residual current. Among them, the leakage currents of each user are difficult to measure, and the change of users' electricity consumption behaviors often affects the magnitudes of both the user load currents and the leakage currents simultaneously. Moreover, the leakage current is a part of the user load current. Therefore, the residual current on the distribution transformer side of the distribution area can be used as the explained variable, the load currents of each user as the explanatory variables, and the leakage current of the supply line as the constant term to construct a multiple linear regression equation for the currents in the distribution area:

[0077]

[0078] Among them, is the phasor value of the residual current, are the phasor values of the load currents of each user, represents the magnitude of the complex weight coefficient of each user's load current with respect to the residual current. N is the number of users supplied by the feeder, is the phasor value of the leakage current of the supply line, represents the magnitude of the leakage current of the i-th user in the distribution area.

[0079] During normal electricity consumption, there is a certain leakage current for each user in the distribution area, which comes from the leakage current of the indoor lines and equipment of the user. However, this part of the leakage current is only a very small part of the load current, and its amplitude is obviously not greater than the load current. When a wiring error fault occurs, the load current will all return to the neutral point of the distribution transformer of the distribution area through the ground wire and be converted into the residual current. Therefore, it can be considered that under any circumstances, the amplitude of the complex weight coefficient of each user's load current with respect to the residual current in the distribution area should not be greater than 1. Thus, the constraint condition for the complex weight coefficient of each user's load current with respect to the residual current in the distribution area can be set as:

[0080]

[0081] Sample the current phasor data at second-level intervals within an interval with a time span of M to obtain the data sets of the residual current in the distribution area and the phasor values of the load currents of each user during this time interval. For each user, construct an auxiliary multiple linear regression model of its load current with respect to the load currents of other users, and calculate the coefficient of determination of the auxiliary multiple linear regression model according to the least squares theory, and further obtain the variance inflation factor of each user. If the variance inflation factors of each user are all less than 10, then each user is linearly independent and the optimal solution of this multiple linear regression equation is unique. If the calculated variance inflation factor of a user is greater than 10, then the number of sample points needs to be further increased until the variance inflation factors of each user are all less than 10.

[0082] The auxiliary multiple linear regression equation is not listed because it needs to be reconstructed for each independent variable. The equation form is still Formula 1. However, the auxiliary multiple linear regression equation means taking the independent variable X1 as the dependent variable, and then taking other X2 to Xn as the independent variables of X1 to construct a multiple regression equation as shown in Equation 1. Its coefficient of determination is calculated based on the auxiliary multiple regression equation. The specific calculation process is shown in the figure. The variance inflation factor is 1 / (1 - coefficient of determination). The calculation of the variance inflation factor is prior art.

[0083] According to Equation 1, construct a system of multiple linear regression equations, and use the least squares theory to calculate the residual sum of squares function value of the residual current sampling value and the estimated value, and make it the smallest to estimate the complex weight coefficient phasor value of each user load current with respect to the residual current in the multiple linear regression equation. Its formula is:

[0084]

[0085] where It can be seen from the residual sum of squares calculation formula that this objective function represents a real-valued complex variable function that maps from the complex domain to the real domain. According to the least squares theory, the objective problem is transformed from a multiple linear regression problem of Equation 1 into a convex function linear optimization problem.

[0086] Using the method of minimizing the residual sum of squares to calculate the optimal regression coefficients of the multiple regression equation belongs to the analytical solution method, and it is difficult to add constraint conditions. Therefore, it is transformed into using the primal-dual interior point method to add constraint conditions and iteratively solve the transformed second-order cone programming problem. Thus, the optimal regression coefficient solution of the multiple linear regression equation that satisfies the constraint conditions is obtained.

[0087] Write Equation 3 in matrix form:

[0088]

[0089] Ignoring the constant part and considering the constraints, its expression is rewritten as the second-order cone programming problem SOCP:

[0090]

[0091] In the formula I r ∈C M×1 , represents M sampling values of the residual current in the distribution transformer area, I L ∈C M×N , represents the sampling values of N user load currents at M moments.

[0092] Furthermore, the primal-dual interior point method is adopted to solve the above SOCP problem. First, the Lagrangian function is constructed. For each constraint, a dual variable λ is introduced to construct the Lagrangian function, which is the sum of the original objective function and all constraint functions (linearly combined by the dual variables). The formula is as follows:

[0093]

[0094] Furthermore, a first-order perturbation KKT condition linear system is constructed, including the primal feasibility condition, the dual feasibility condition, and the complementary slackness condition. The formulas are as follows:

[0095]

[0096] In the formulas, denotes taking partial derivatives of the variables in L(β,λ 1 ,...,λ N ) one by one. Seek partial derivatives one by one.

[0097] is a 3*N matrix. μ is a positive decimal, which gradually decreases during the iteration process to ensure that the entire iteration process is always in the interior point region.

[0098] Furthermore, the Newton iteration method is used to solve the KKT condition equation to find the optimal solution along the central path.

[0099] The values of (β,λ) are updated using the following formula:

[0100]

[0101] The updated β k+1 and λ k+1 can be used to update the value of μ k+1 :

[0102] μ k+1 =σ(β k+1 ) T λ k+1 / N (9)

[0103] Among them, the primal step size α_p and the dual step size α_d are calculated using the following formula.

[0104]

[0105] The iteration direction of the Newton method is (Δβ k ,Δλ k ), where both Δβ k and Δλ k are N*1 matrices:

[0106]

[0107] Δβ k , Δλ k is calculated using the following formula:

[0108]

[0109] The initial values are (β 0 , λ 0 ), satisfying β 0 > 0, λ 0 > 0, |β 0 | 2 ≤ 1,, where β 0 is the original variable, λ 0 is the dual variable. At the same time, a positive number μ 0 and a parameter σ between 0 and 1 are selected. Each element in β 0 , each element in λ 0 , μ 0 , and σ are all preset values. The value ranges of μ 0 and σ are both (0, 1]. Each element in β 0 is equal and the value range is (0, 1]. Each element in λ 0 is equal and the value range is (0, 1]. In this embodiment, each element in β 0 takes the value of 1..., and each element in λ 0 takes the value of 0.1, the value of μ 0 is 0.1, and the value of σ is 0.1. represents taking the partial derivative of with respect to in . That is, first regard as a variable, take the partial derivative, and then substitute the obtained and the obtained into the expression after taking the partial derivative.

[0110] J(β k , λ k ) represents the Jacobian matrix of F(β k , λ k ):

[0111]

[0112] The number of rows of J(β k , λ k ) is 3N, and the number of columns is 2N, where:

[0113]

[0114] where q = 1, 2, 3; F q1 , F q2, ……, F qN respectively correspond to and represent the element in the first column of the q-th row, the element in the second column of the q-th row, ……, the element in the N-th column of the q-th row in the matrix F(β k , λ k ). That is, in each iteration, each complex weight coefficient is updated.

[0115]

[0116]

[0117] where N is the dimension of the variable. Further calculate the residual r(β, λ) k+1 =(r_c k+1 , r_b k+1 , r_s k+1 ), where:[[]]

[0118]

[0119] Here, the value of β is continuously calculated and updated through formulas (10)-(12), and further, whether the magnitude of the two-norm of the residual r meets the requirement for ending the iteration is obtained through formulas (13) and (14) after each iterative update. r_c k+1 , r_b k+1 , r_s k+1 are all 1×N matrices (i.e., N-dimensional vectors), and the three are combined to obtain the 1×3N matrix r(β, λ) k+1 . Whether to end the iteration is judged by the value of the two-norm of the residual r(β, λ) k+1 , that is, the value obtained by taking the square root of the sum of the squares of the elements in the 1*3N vector.

[0120] If the number of iterations reaches the preset number of iteration values, or the two-norm of the residual r(β, λ) k+1 is less than or equal to the preset norm threshold, then the iteration stop condition is satisfied, the iteration is stopped, and β k+1 is used as the final complex weight coefficient phasor Otherwise, continue the iteration.

[0121] By iteratively updating the step size and search direction and calculating the residuals of the primal problem and the dual problem, if the norm of the residual r(β, λ) k+1 is small enough, or the preset number of iterations or time limit is reached, then the iteration is stopped, and the current solution is output as the optimal solution found within the feasible region under the preset conditions. The value range of the preset norm threshold is 10 -3 or 10 -4 .

[0122] In practical problems, the obtained residual current phasor data I of the transformer substation area can be usedr and the load current data I of each user L are input into Matlab to form a Mat file. Using the CVX package function, considering the above constraints, the optimal complex weight coefficient phasor that minimizes E(β) is calculated within the complex feasible region, and is used as an index to measure the strength of the correlation between the load current of each user and the residual current in the distribution transformer area. The example shows that the proposed method for identifying the wiring error and leakage fault users based on the primal-dual interior point method can accurately distinguish the existence of abnormal users and make the calculation results more interpretable.

[0123] It should be noted that the embodiments in this specification are all described in a progressive manner. Each embodiment focuses on the differences from other embodiments. For the same and similar parts among the embodiments, reference can be made to each other.

[0124] The above has described the embodiments of the present invention in detail, but the content described is only the preferred embodiments of the present invention and cannot be considered as limiting the scope of implementation of the present invention. All equivalent changes and improvements made according to the scope of the present invention should still fall within the scope covered by the present invention. After reading the present invention, various equivalent forms of modification by those skilled in the art fall within the scope defined by the appended claims of the present invention. Without conflict, the embodiments in the present invention and the features in the embodiments can be combined with each other.

Claims

1. A method for detecting users with wiring errors and leakage, characterized in that, it includes: (A) During the acquisition time period, the residual current of the measured substation area and the load currents of each user in the measured substation area are acquired at a set sampling interval; a residual function is constructed, and the residual function is the sum of the squared residuals of the sampled values and the estimated values of the residual current of the measured substation area during the acquisition time period; (B) By minimizing the value of the residual function, the complex weight coefficients of each user's load current with respect to the residual current of the measured substation area in the multiple linear regression equation are estimated; (C) If the amplitude of the complex weight coefficient of the j-th user obtained by estimation is greater than or equal to the preset amplitude threshold, it is determined that the j-th user is a user with wiring errors and leakage; otherwise, it is determined that the j-th user is a normal user; where j ∈ [1, N], and N represents the number of users in the measured substation area.

2. The method for detecting users with wiring errors and leakage according to claim 1, characterized in that, the constructed residual function is: wherein, is the complex weight coefficient of the load current of the j-th user with respect to the residual current of the measured power distribution area; is the i-th sampling value of the residual current of the measured power distribution area, is the i-th sampling value of the load current of the j-th user; i ∈ [1, M], M represents the number of sampling values of the residual current of the measured power distribution area; M > N.

3. The method for detecting users with wiring errors and leakage according to claim 1, characterized in that, in the step (B), the specific method for estimating the complex weight coefficients of each user's load current with respect to the residual current of the measured substation area by minimizing the value of the residual function is: Solve the constructed second-order cone programming problem to obtain the complex weight coefficient of the load current of the j-th user in the measured power distribution area with respect to the residual current of the measured power distribution area value; The constructed second-order cone programming problem is: under the condition of satisfying , minimize the value of E'(β); Among them, Residual current matrix represents the conjugate transpose matrix of I r , represents the conjugate transpose matrix of I L , β H represents the conjugate transpose matrix of β, load current matrix Among them, is the complex weight coefficient of the load current of the j-th user with respect to the residual current of the measured power distribution area; is the i-th sampling value of the residual current of the measured power distribution area, is the i-th sampling value of the load current of the j-th user; i ∈ [1, M], where M represents the number of sampling values of the residual current of the measured power distribution area; M > N.

4. The method for detecting users with wiring errors and leakage according to claim 3, characterized in that, the quadratic cone programming problem constructed is solved by using the primal-dual interior point method.

5. The method for detecting users with wiring errors and leakage according to claim 4, characterized in that, the specific method for solving the quadratic cone programming problem constructed by using the primal-dual interior point method is: (B1) The following formula is used for iterative operation: where: J(β k ,λ k ) represents the Jacobian matrix of F(β k ,λ k ); denotes taking the partial derivative with respect to in . Among them, is the constructed Lagrangian function, (β k ) H represents the conjugate transpose matrix of β k , represents 's amplitude, represents the conjugate transpose matrix of I L , represents the conjugate transpose matrix of I r ; α_p is the original step size, α_d is the dual step size, β k is the k-th complex weight coefficient update matrix; λ k is the k-th dual variable matrix; Δβ k is the k-th first increment matrix, Δλ k is the k-th second increment matrix; k = 0, 1, ……; Each element in β 0 , each element in λ 0 , μ 0 , and σ are all preset values. The value ranges of μ 0 and σ are both (0, 1]. Each element in β 0 is equal and its value range is (0, 1]. Each element in λ 0 is equal and its value range is (0, 1]; (B2) Determine whether the iteration stop condition is satisfied. If the judgment result is yes, stop the iteration and use β k+1 as the obtained complex weight coefficient β. Otherwise, continue the iteration.

6. The method for detecting users with wiring errors and leakage according to claim 5, characterized in that, where q = 1, 2, 3; F q1 , F q2 , ……, F qN respectively correspond to the elements in the first column of the q-th row, the elements in the second column of the q-th row, ……, the elements in the N-th column of the q-th row in the matrix F(β k , λ k ).

7. The method for detecting users with wiring errors and leakage according to claim 5, characterized in that, the iteration stop condition is: the number of iterations reaches the preset number of iteration values; or The two - norm of the residual r(β,λ) k+1 is less than or equal to a preset norm threshold; where the residual \(r(\beta,\lambda)\) k+1 =(r_c k+1 ,r_b k+1 ,r_s k+1 ); 8. The method for detecting users with wiring errors and leakage according to claim 7, characterized in that, The value range of the preset norm threshold is [10 -4 , 10 -3 .

9. The method for detecting users with wiring errors and leakage according to claim 1, characterized in that, the value range of the preset amplitude threshold is [0.5, 0.9].

10. A device for detecting users with wiring errors and leakage, characterized in that, it includes a computer device; the computer device is configured or programmed to execute the steps of the method for detecting users with wiring errors and leakage according to any one of claims 1-9.

Images