Electricity stealing identification method and device based on power distribution network state estimation

Abstract

The invention discloses an electricity stealing identification method and device based on state estimation of a power distribution network. The method comprises the following steps: acquiring measurement data from a D5000 system, an electricity acquisition terminal message and an electricity utilization information acquisition system; performing state estimation calculation on the line according to the measured values of all positions of the line node obtained through calculation; the method comprises the following steps of: performing arrangement according to measurement items, reducing or zeroing a to-be-detected measurement data weight by adopting a weight reduction method, recalculating a line operation state by adopting the measurement data after weight reduction, and repeating the process until all measurement is subjected to weight reduction and recalculation respectively. And solving a residual error of a weight reduction result of each time and calculating a correction rate of the residual error so as to judge whether abnormal electricity utilization exists in the measured value node or not, thereby realizing the purpose of executing electricity larceny identification by the electricity larceny identification method and device based on the state estimation of the power distribution network.

Description

Technical Field

[0001] The present invention belongs to the technical field of electricity theft identification, and in particular relates to an electricity theft identification method and device based on distribution network state estimation. Background Art

[0002] Electricity theft is the illegal occupation of electrical energy with the purpose of not paying or paying less electricity bills, using illegal means to not measure or under-measure electricity usage.

[0003] For illegal acts such as electricity theft, corresponding detection and identification must be performed. At present, power supply companies use traditional manual censuses as mentioned in the existing technical solutions with patent application number "CN202310952325.9" and patent name "Anti-electricity theft method, system, equipment and storage medium" to detect electricity theft. The workload is large, the pertinence is insufficient, and it is easy to alert the enemy; it is difficult to control the cost by installing segmented metering; and the correlation coefficient method evaluates the user's suspicion of electricity theft by calculating the correlation coefficient between the user's electricity consumption and line loss electricity, and the user's electricity consumption and power supply, and cannot determine the zero-electricity theft user; another method based on user electricity consumption feature cluster analysis can identify the conversion of user electricity consumption mode, but cannot effectively identify abnormal electricity consumption behaviors such as long-term electricity theft by users. Therefore, a universal algorithm that goes beyond the analysis of electricity consumption for a single user is needed to accurately detect electricity theft, that is, a method and device for identifying electricity theft based on the state estimation of the distribution network is needed to perform electricity theft identification, but there is currently no technology for identifying electricity theft based on the state estimation of the distribution network. Summary of the invention

[0004] In order to solve the defects in the prior art, the present invention proposes a method and device for identifying electricity theft based on the state estimation of the distribution network, which obtains measurement data from the D5000 system, the power acquisition terminal message, and the power consumption information collection system; the state estimation calculation of the line is performed according to the measurement values ​​at each line node; the measurement items are arranged according to the measurement items, and the weight reduction method is used to reduce or set the weight of a measurement data to be detected, and the line operation state is recalculated using the weighted measurement data, and this process is repeated until all measurement data are weighted and recalculated. The residual is obtained for each weight reduction result and its correction rate is calculated to determine whether the measurement value node has abnormal electricity consumption, thereby achieving the purpose of performing electricity theft identification based on the distribution network state estimation method and device.

[0005] The present invention uses the following technical solutions.

[0006] A method for identifying electricity theft based on distribution network state estimation, comprising:

[0007] Step 1: Obtain measurement data from the D5000 system, power acquisition terminal messages, and power consumption information collection system;

[0008] Step 2: Perform state estimation calculation on the line based on the calculated measurement values ​​at each line node;

[0009] Step 3: First, arrange the measurement items, use the weight reduction method to reduce or set the weight of a measurement data to be detected, and use the reduced-weight measurement data to recalculate the line operation status. Repeat this process until all measurements are reduced-weighted and recalculated. For each reduced-weight result, obtain the residual and calculate its correction rate to determine whether there is abnormal power consumption at the measurement value node.

[0010] Preferably, if the measured data in step 1 lacks relevant values ​​or contains irregular data, an interpolation algorithm is used to supplement the missing relevant values ​​or a transformation method is used to transform the irregular data into regular data.

[0011] Preferably, step 1 specifically includes:

[0012] For line junctions, there is phase voltage telemetry data Specifically including A phase voltage telemetry data Phase B voltage telemetry data Phase C voltage telemetry data It can be obtained from the D5000 system; the high-voltage power meter at the gateway measures the line voltage between phase A and phase B Line voltage between phase C and phase B Line voltage between phase A and phase C Phase A current Phase B current Phase C current The total active power P, total reactive power Q and voltage phase angle information of each phase, the data measured by the high-power supply and high-power meter at the gateway are stored in the remote energy collection terminal in the station as the terminal message of the electric energy collection terminal, and can be obtained through the master station call measurement connected to the remote energy collection terminal in the station. The measurement data includes the phase voltage telemetry data of the line gateway Line voltage between phase A and phase B Line voltage between phase C and phase B Line voltage between phase A and phase C Phase A current Phase B current Phase C current Total active power P, total reactive power Q and voltage phase angle information of each phase.

[0013] Preferably, step 1 specifically further includes: for high-supply and high-metering users, the electric energy meter used as the power consumption information collection system measures the line voltage Phase current And the total active power Pi , total reactive power Q i ; Then supplement the AC phase line voltage A, B, C phase voltage Phase B current And related phase angle information, the supplementary methods include:

[0014] The three-phase voltage phase angle calculated at the line junction is used To compensate for the unknown load three-phase voltage phase angle, the voltage calculated at the line junction is used. Compensating for unknown loads

[0015] Then, according to the three-variable quadratic equation system shown in formula (8):

[0016]

[0017] Available The measurement data also includes the line voltage for high-power and high-meter users Phase current And the total active power P i , total reactive power Q i , AC phase voltage A, B, C phase voltage Phase B current And the related voltage phase angle information.

[0018] Preferably, step 1 specifically also includes: for the transformer branch with high supply and low metering, and the nameplate data that has been converted to the high voltage side, an equivalent mathematical model is established, and the taps of the equivalent mathematical model are all put into corresponding gears. After the measured data is transformed, the solution of branch i-j' is equivalently converted to the solution of branch ij.

[0019] Preferably, step 1 specifically further includes: adding a set of pseudo-measured values ​​of iron loss as node injection power on one side of the ideal transformer, as shown in formula (9):

[0020]

[0021] In the formula, ΔS T0 is the rated loss of the transformer at no load; U av is the average rated voltage of the phase; U tap is the transformer tap voltage, ΔA0 is the power injected into the node on one side of the ideal transformer, and the measured data also includes the pseudo-measurement value of iron loss.

[0022] Preferably, in step 1, the measurement data also includes virtual measurement values.

[0023] Preferably, in step 1, the measurement data further includes core quantity measurement.

[0024] Preferably, step 2 specifically includes:

[0025] The relationship between the measurement data and the state variables is expressed by the following measurement equation, namely formula (10):

[0026] z=h(x)+e (10)

[0027] Where z∈R m is the measurement vector; e=[e1,e2,……,e m ] T ∈R m is the measurement error vector; x is the state variable, which is a set of independent, minimal variables that can characterize all the characteristics of the power system. Let x = [θ T ,v T ] includes all node voltage magnitudes and phase angles (except the reference node phase angle), θ∈R N-1 is the phase angle of the node except the reference node, v∈R N is the voltage modulus of all nodes. After determining the voltage modulus and phase angle, the operating state of the entire system can be determined accordingly; h:R n →R m is the nonlinear mapping from state variables to measurement vectors, i.e., the measurement expression;

[0028] The voltage measurement expression is shown in formula (11):

[0029] v i =v i (11)

[0030] The node injection power measurement expression is shown in formula (12):

[0031]

[0032] In formula (11) and formula (12), v i and θ i are the voltage magnitude and phase angle of node i respectively; θ ij =θ i -θ j ;

[0033] The residual vector is expressed as shown in formula (13):

[0034]

[0035] In the formula, is the estimated value of the state variable vector;

[0036] In order to make the estimated line operation status fit the measured value best, based on formula (10) and the least squares method, the following mathematical model is established and solved, that is, formula (14):

[0037]

[0038] The partial derivative of the objective function J(x) of formula (14) is calculated and set to 0, as shown in formula (15):

[0039]

[0040] In the formula, It is called the Jacobian matrix;

[0041] Construct H(x) using formula (16):

[0042]

[0043] Where 0 represents a vector or matrix with all elements equal to 0; θ and v represent the phase angle vector and the voltage magnitude vector of all nodes except the reference node, respectively; v m represents the node voltage modulus measurement vector; P inj and Q inj They represent the injected active power measurement and injected reactive power measurement of the node respectively.

[0044] Preferably, the specific expression of H(x) is as follows:

[0045] Use formula (17) to obtain the partial derivative of the node voltage measurement for the phase angle and voltage of the node:

[0046]

[0047] Use formula (18) to obtain the partial derivative of the node voltage measurement with respect to the phase angle and voltage of the adjacent node:

[0048]

[0049] Use formula (19) to measure the active power injection of the node and obtain the partial derivative of the phase angle and voltage of the node:

[0050]

[0051] The active power injection measurement of the node is used to obtain the partial derivative of the phase angle and voltage of the adjacent node:

[0052]

[0053] The reactive power injection measurement of the node is used to obtain the partial derivative of the phase angle and voltage of the node:

[0054]

[0055] The reactive power injection measurement of the node is used to obtain the partial derivative of the phase angle and voltage of the adjacent node:

[0056]

[0057] Let the equivalent first-order optimal condition satisfied by formula (14) be g(x), which is shown in formula (23):

[0058] g(x)=-H T (x)[zh(x)] (23)

[0059] Put g(x) at x (k) Perform an approximate Taylor series expansion at to linearize it, and we can get formula (24):

[0060]

[0061] Let g′(x)=G(x), and we get the linear form of the canonical equation, which is shown in formula (25):

[0062] G(x (k) )Δx (k) =H T (x (k) )[zh(x (k) )] (25)

[0063] Where k is the number of iterations; Δx (k) =x (k+1) -x (k) ; G=H T H is the gain matrix, which is a symmetric sparse positive definite matrix; H T (x (k) )[zh(x (k) )] is the free vector;

[0064] Modify the state variable x (2) =x (1) +Δx (1) , get the state variable x after the first iteration correction (2) , continue to iterate, that is, perform iterative correction according to formula (25) until the objective function J(x (k) ) until the convergence criterion is reached, which is shown in formula (26):

[0065] ||Δx (k) ||≤ε (26)

[0066] In the formula, ||·|| represents the infinite norm; ε is the convergence criterion, at which point the estimated values ​​of all nodes can be obtained

[0067] A power theft identification device based on distribution network state estimation, comprising:

[0068] An acquisition module is used to obtain measurement data from the D5000 system, power acquisition terminal messages, and power consumption information acquisition system;

[0069] An estimation module, which is used to perform state estimation calculation on the line according to the calculated measurement values ​​at each line node;

[0070] The judgment module is used to arrange the measurement items, use the weight reduction method to reduce or set the weight of a measurement data to be detected, and recalculate the line operation status using the weighted measurement data. This process is repeated until all measurement items are weighted down and recalculated. The residual is obtained for each weight reduction result and its correction rate is calculated to determine whether there is abnormal power consumption at the measurement value node.

[0071] The beneficial effect of the present invention is that, compared with the prior art, the present invention obtains measurement data from the D5000 system, the power acquisition terminal message, and the power consumption information collection system; performs state estimation calculations on the line according to the calculated measurement values ​​at each line node; arranges them according to the measurement items, uses the weight reduction method to reduce or set the weight of a measurement data to be detected, and uses the weighted measurement data to recalculate the line operation state, and repeats this process until all measurement measurements are weighted and recalculated. The residual is obtained for each weight reduction result and its correction rate is calculated to determine whether the measurement value node has abnormal power consumption, thereby achieving the purpose of performing power theft identification based on the power distribution network state estimation method and device to perform power theft identification. BRIEF DESCRIPTION OF THE DRAWINGS

[0072] Figure 1 It is an overall flow chart of the method for identifying electricity theft based on distribution network state estimation described in the present invention;

[0073] Figure 2 It is a schematic diagram of the module structure of the power theft identification device based on distribution network state estimation in the present invention;

[0074] Figure 3 It is the phase quantity diagram of the high-power supply and high-power meter electric energy meter mentioned in the present invention;

[0075] Figure 4 It is a structural diagram of the high supply and low meter transformer branch equivalent model described in the present invention. DETAILED DESCRIPTION

[0076] Manual surveys to detect electricity theft are labor-intensive, lack pertinence, and are easy to alert the enemy; it is difficult to control costs by installing segmented metering; and the correlation coefficient method evaluates the user's suspicion of electricity theft by calculating the correlation coefficient between the user's electricity consumption and line loss electricity, and the user's electricity consumption and power supply, but it is unable to identify zero-power electricity theft users; other methods based on user electricity consumption characteristics cluster analysis can identify the conversion of user electricity consumption patterns, but cannot effectively identify abnormal electricity consumption behaviors such as long-term electricity theft by users. Therefore, a universal algorithm that goes beyond the analysis of electricity consumption for a single user is needed to accurately detect electricity theft, that is, a method and device for identifying electricity theft based on distribution network state estimation is needed to perform electricity theft identification, but there is currently no technology for identifying electricity theft based on distribution network state estimation. Therefore, the technical solution of the present invention is proposed to solve the above problems.

[0077] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely expressed in combination with the drawings in the embodiments of the present invention. The embodiments expressed in this application are only part of the embodiments of the present invention, not all of them. According to the spirit of the present invention, other embodiments obtained by ordinary technicians in this field without creative work are all within the scope of protection of the present invention.

[0078] like Figure 1 As shown, the method for identifying electricity theft based on distribution network state estimation according to the present invention comprises:

[0079] Note that electricity theft only changes the metering value of the electricity theft user, but does not change the fact that the electricity theft user is still using electricity, nor does it change the power distribution of the distribution network. Therefore, the line state closest to the current measurement is calculated through the measurement and metering values ​​at various points of the line, and the place with a large difference from the metering value is the electricity theft user.

[0080] Therefore, the present invention has a method for identifying electricity theft based on distribution network state estimation, which specifically includes:

[0081] Step 1: Obtain measurement data from the D5000 system (gateway telemetry value), power acquisition terminal message (gateway metering value), and power consumption information collection system (load metering value);

[0082] In a preferred but non-limiting embodiment of the present invention, if the measured data in step 1 lacks relevant values ​​or has irregular data, an interpolation algorithm is used to supplement the missing relevant values ​​or a transformation method is used to transform the irregular data into regular data. The interpolation algorithm and the transformation method are both prior art interpolation algorithms and transformation methods in data cleaning.

[0083] Some data sets lack relevant values, and some cannot be directly involved in the calculation and need to be transformed and supplemented through mathematical methods to meet the calculation requirements.

[0084] In a preferred but non-limiting embodiment of the present invention, Figure 3 From the high supply and high metering energy meter phase diagram, we can see that step 1 specifically includes:

[0085] For line junctions, there is phase voltage telemetry data Specifically including A phase voltage telemetry data Phase B voltage telemetry data Phase C voltage telemetry data It can be obtained from the D5000 system; the high-voltage power meter at the gateway measures the line voltage between phase A and phase B Line voltage between phase C and phase B Line voltage between phase A and phase C Phase A current Phase B current Phase C current The total active power P, total reactive power Q and voltage phase angle information related to each phase, the data measured by the high-power supply and high-power meter at the gateway are stored in the remote energy collection terminal in the station as the terminal message of the electric energy collection terminal, and can be obtained through the master station call measurement connected to the remote energy collection terminal in the station. The measurement data includes the phase voltage telemetry data for the line gateway Line voltage between phase A and phase B Line voltage between phase C and phase B Line voltage between phase A and phase C Phase A current Phase B current Phase C current Total active power P, total reactive power Q and voltage phase angle information related to each phase.

[0086] In some scenarios, the AC phase line voltage is missing and needs to be supplemented Phase B current And the related phase angle information, the supplementary method is calculated as formula (1) to formula (6):

[0087]

[0088]

[0089] Then we can obtain α1 and α2

[0090]

[0091]

[0092] It can be obtained from this

[0093]

[0094] The total active power and total reactive power measured by the gateway electric energy meter are

[0095]

[0096] Formula (6) is a set of two-variable transcendental equations, which can produce two sets of solutions. Negative solution can be obtained θ, so we can get Depend on Can be obtained Based on the voltage and current of each phase, the phase-by-phase active power of the line junction can also be obtained. Phase reactive power

[0097] In a preferred but non-limiting embodiment of the present invention, step 1 specifically further includes: for high-supply high-metering users, the electric energy meter used as the power consumption information collection system measures the line voltage Phase current And the total active power P i , total reactive power Q i ; Compared with the gateway side, there is a lack of phase voltage telemetry data The AC phase voltage needs to be supplemented A, B, C phase voltage Phase B current And related phase angle information, the supplementary method includes: the total active power and total reactive power measured by the electric energy meter are the same as formula (6), but two more table display data are available for use, that is, using formula (7):

[0098]

[0099] Where P1 and P2 are the measured values ​​of phase A and C on the high-supply and high-meter electric energy meter, and do not contain the secondary power with physical meaning. AB ,I A , U CB ,I C The A-phase voltage and current and the C-phase voltage and current measured by the electric energy meter correspond to the AB-phase line voltage modulus and the A-phase current modulus, and the BC-phase line voltage modulus and the C-phase current modulus.

[0100] Among the three-phase imbalance, most of them are three-phase current imbalance, the modulus imbalance of the three-phase voltage is very small, and the voltage phase angle drop is even smaller, because the three-phase voltage drop difference caused by load imbalance is even smaller. [2] Therefore, in the measurement value transformation, the three-phase voltage phase angle calculated at the line junction is used. To compensate for the unknown load three-phase voltage phase angle, the voltage calculated at the line junction is used. Compensating for unknown loads

[0101] Then, according to the three-variable quadratic equation system shown in formula (8):

[0102]

[0103] Available The measurement data also includes the line voltage for high-power and high-meter users Phase current And the total active power P i , total reactive power Q i , AC phase voltage A, B, C phase voltage Phase B current And related voltage phase angle information. Thus, the phase voltage modulus of all high-supply and high-load nodes can be obtained. Phase active power Reactive power

[0104] In a preferred but non-limiting embodiment of the present invention, step 1 specifically includes: for high-supply and low-metering users or substations, due to the influence of the metering method, the measurement data obtained by the power consumption information collection system cannot be directly used in the calculation. Therefore, for the transformer branch with high supply and low metering, and the nameplate data that has been converted to the high voltage side, the extremely small excitation current on R and X is ignored, and the following is established: Figure 4 The equivalent mathematical model shown in Figure 4 Medium F The position of the transformer tap should be consistent with the actual on-site tap position. For example, when the 838 Guorui Line 1 Longhu Tianyi 2# public transformer was put into operation, the taps of the equivalent mathematical model were all put into the corresponding gear (1+2*2.5%), so the actual transformer ratio was 10 / 0.4*(1+2*2.5%). After the measurement data is transformed, the solution of branch i-j' is equivalently transformed into the solution of branch ij.

[0105] In a preferred but non-limiting embodiment of the present invention, step 1 specifically further includes: in order to avoid the excitation branch affecting the generation of the node admittance matrix while not ignoring its loss, a set of pseudo-measured values ​​of iron loss are added on one side of the ideal transformer as the node injection power, as shown in formula (9):

[0106]

[0107] In the formula, ΔS T0 is the rated loss of the transformer at no load; U av is the average rated voltage of the phase; Utap is the transformer tap voltage, ΔA0 is the power injected into the node on one side of the ideal transformer, and is the optimal engineering data calculated to fit the actual operation of the transformer high supply and low metering method. It is a set of fictitious measurements and does not actually exist. It is called "pseudo-measurement values". The measurement data also includes pseudo-measurement values ​​of iron loss.

[0108] In a preferred but non-limiting embodiment of the present invention, in step 1, there are some intermediate nodes, such as nodes characterized by having neither power points nor loads, which are called "zero injection nodes". Such nodes strictly satisfy Kirchhoff's law. Although there is no actual measurement, their input and output power are zero, that is, P i =0, Q i =0, called "virtual measurement value", the measurement data also includes virtual measurement value.

[0109] In a preferred but non-limiting embodiment of the present invention, in step 1, core quantity measurement refers to one or a group of quantity measurements that must be possessed in order for the measurement equation to cover all state variables. After the core quantity measurement is missing, the measurement equation of the entire network cannot cover all state variables. For example, for a branch where only the terminal voltage and power are known, its voltage and power are both core quantity measurements, and neither is missing. The measurement data also includes the core quantity measurement.

[0110] Step 2: Perform state estimation calculation on the line based on the calculated measurement values ​​at each line node;

[0111] In a preferred but non-limiting embodiment of the present invention, step 2 specifically comprises:

[0112] The relationship between the measurement data and the state variables is expressed by the following measurement equation, namely formula (10):

[0113] z=h(x)+e (10)

[0114] Where z∈R m is the measurement vector (a vector composed of measurement data); e = [e1, e2, ..., e m ] T ∈R m is the measurement error vector; x is the state variable, which is a set of independent, minimal variables that can characterize all the characteristics of the power system. Let x = [θ T ,v T ] includes all node voltage magnitudes and phase angles (except the reference node phase angle), θ∈R N-1 is the phase angle of the node except the reference node, v∈R N is the voltage modulus of all nodes. After determining the voltage modulus and phase angle, the operating state of the entire system can be determined accordingly; h:R n →Rm is the nonlinear mapping from state variables to measurement vectors, i.e., the measurement expression;

[0115] The voltage measurement expression is shown in formula (11):

[0116] v i =v i (11)

[0117] The node injection power measurement expression is shown in formula (12):

[0118]

[0119] In formula (11) and formula (12), v i and θ i are the voltage magnitude and phase angle of node i respectively; θ ij =θ i -θ j ;

[0120] The difference between the measured value and the estimated value is called the residual. The residual vector is expressed as shown in formula (13):

[0121]

[0122] In the formula, is the estimated value of the state variable vector; ideally, each component of the residual vector should be 0. However, since the dimension of the quantity measurement is greater than the dimension of the state variable, and there are actually random errors in the quantity measurement, the equation system is an unsolvable contradictory equation system, that is, there is no Make all components of the residual vector equal to 0.

[0123] In order to make the estimated line operation status fit the measured value best, based on formula (10) and the least squares method, the following mathematical model is established and solved, that is, formula (14):

[0124]

[0125] The single-phase voltage of 6kV is used as the reference value for level start, and the initial value of the state variable of each node is x (1) As expressed by the value set in advance.

[0126] In order to make the objective function J(x) of formula (14) obtain the optimal value, the partial derivative of the objective function J(x) of formula (14) is calculated and set to 0, as shown in formula (15):

[0127]

[0128] In the formula, It is called the Jacobian matrix;

[0129] Construct H(x) using formula (16):

[0130]

[0131] Where 0 represents a vector or matrix with all elements equal to 0; θ and v represent the phase angle vector and the voltage magnitude vector of all nodes except the reference node, respectively; v m represents the node voltage modulus measurement vector; P inj and Q inj They represent the injected active power measurement and injected reactive power measurement of the node respectively.

[0132] In a preferred but non-limiting embodiment of the present invention, the specific expression of H(x) is as follows:

[0133] Use formula (17) to obtain the partial derivative of the node voltage measurement for the phase angle and voltage of the node:

[0134]

[0135] Use formula (18) to obtain the partial derivative of the node voltage measurement with respect to the phase angle and voltage of the adjacent node:

[0136]

[0137] Use formula (19) to measure the active power injection of the node and obtain the partial derivative of the phase angle and voltage of the node:

[0138]

[0139] The active power injection measurement of the node is used to obtain the partial derivative of the phase angle and voltage of the adjacent node:

[0140]

[0141] The reactive power injection measurement of the node is used to obtain the partial derivative of the phase angle and voltage of the node:

[0142]

[0143] The reactive power injection measurement of the node is used to obtain the partial derivative of the phase angle and voltage of the adjacent node:

[0144]

[0145] Let the equivalent first-order optimal condition satisfied by formula (14) be g(x), which is shown in formula (23):

[0146] g(x)=-H T (x)[zh(x)] (23)

[0147] For power systems, g(x) is a set of n-dimensional nonlinear algebraic equations. (k) Perform an approximate Taylor series expansion at to linearize it, and we can get formula (24):

[0148]

[0149] Let g′(x) = G(x), ignore the quadratic and higher-order terms in equation (2-24), and we can get the linear form of the regular equation, which is shown in equation (25):

[0150] G(x (k) )Δx (k) =H T (x (k) )[zh(x (k) )] (25)

[0151] Where k is the number of iterations; Δx (k) =x (k+1) -x (k) ; G=H T H is the gain matrix, which is a symmetric sparse positive definite matrix; H T (x (k) )[zh(x (k) )] is the free vector;

[0152] Obtain Therefore, the state variable x is modified (2) =x (1) +Δx (1) , get the state variable x after the first iteration correction (2) , continue to iterate, that is, perform iterative correction according to formula (25) until the objective function J(x (k) ) until the convergence criterion is reached. The commonly used convergence criterion is shown in formula (26):

[0153] ||Δx (k) ||≤ε (26)

[0154] In the formula, ||·|| represents the infinite norm; ε is the convergence criterion, which is generally taken as 10 for 10kV voltage level. -5 At this point, the estimated values ​​of all nodes can be obtained

[0155] Step 3: To determine the specific abnormal power consumption node, first arrange the measurement items according to the measurement items, use the weight reduction method to reduce or set the weight of a measurement data to be detected (active and reactive power appear in pairs, so the weight is reduced in pairs), and use the weighted measurement data to recalculate the line operation status. Repeat this process until all measurements (virtual measurements, pseudo measurements, and measurements at the gateway are considered to be normal and do not participate) are weighted and recalculated. For each weight reduction result, the residual is calculated and its correction rate is calculated, that is, the difference ratio between the new estimated result after the measurement is reduced and the original measurement value, so as to determine whether the measurement value node has abnormal power consumption. If the difference ratio is higher than the preset threshold, there is abnormal power consumption, otherwise there is no abnormal power consumption.

[0156] In the gain matrix G = H T H and free vector H T (x (k) )[zh(x (k) )], a residual calculation unit diagonal matrix C = diag{1,1,1,...,1} is added. When searching for residuals, the corresponding elements of the non-core measurement values ​​to be identified in C can be directly set to 0 to simplify programming, or the corresponding elements of the core measurement values ​​in C can be set to 10. -3 ~10 -4 Reduce the weight.

[0157] Then the gain matrix G for residual detection is H T CH, the free vector used for residual detection is H T (x (k) )C[zh(x (k) )]. For each downgraded result, obtain the residual and calculate its correction rate.

[0158] The method of the present invention has been applied to locate and find users who steal electricity from distribution lines. The verification results include but are not limited to line-transformer relationship errors, on-site meter flight, low-voltage users connecting to public transformers without meters above the transformer, and high-supply and low-recording transformers short-circuiting currents K1K2.

[0159] like Figure 2 As shown, the power theft identification device based on distribution network state estimation of the present invention comprises:

[0160] An acquisition module is used to obtain measurement data from the D5000 system, power acquisition terminal messages, and power consumption information acquisition system;

[0161] An estimation module, which is used to perform state estimation calculation on the line according to the calculated measurement values ​​at each line node;

[0162] The judgment module is used to arrange the measurement items, use the weight reduction method to reduce or set the weight of a measurement data to be detected, and recalculate the line operation status using the weighted measurement data. This process is repeated until all measurement items are weighted down and recalculated. The residual is obtained for each weight reduction result and its correction rate is calculated to determine whether there is abnormal power consumption at the measurement value node.

[0163] The beneficial effect of the present invention is that, compared with the prior art, the present invention obtains measurement data from the D5000 system, the power acquisition terminal message, and the power consumption information collection system; performs state estimation calculations on the line according to the calculated measurement values ​​at each line node; arranges them according to the measurement items, uses the weight reduction method to reduce or set the weight of a measurement data to be detected, and uses the weighted measurement data to recalculate the line operation state, and repeats this process until all measurement measurements are weighted and recalculated. The residual is obtained for each weight reduction result and its correction rate is calculated to determine whether the measurement value node has abnormal power consumption, thereby achieving the purpose of performing power theft identification based on the power distribution network state estimation method and device to perform power theft identification.

[0164] The present disclosure can be a system, method and / or computer program product. The computer program product can include a computer-readable backup medium having computer-readable program instructions thereon for causing a processor to execute various aspects of the disclosure.

[0165] The computer readable backup medium can be a tangible network capable of retaining and backing up the instructions used by the instruction execution network. The computer readable backup medium can be, but is not limited to, an electrical backup network, a magnetic backup network, an optical backup network, an electromagnetic backup network, a semiconductor backup network, or any suitable combination thereof. Further examples of computer readable backup media (non-exhaustive list) include: portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), static random access memory (SRAM), portable compact disk read-only memory (HD-ROM), digital versatile disk (DXD), memory stick, floppy disk, mechanically encoded network circuits, such as punch cards or protrusions in grooves with instructions backed up thereon, and any suitable combination thereof. The computer readable backup medium used herein is not to be construed as a transient signal per se, such as a radio wave or other freely propagating electromagnetic wave, an electromagnetic wave propagating via a waveguide or other transmission medium (e.g., a light pulse via a power line cable), or an electrical signal transmitted via wires.

[0166] The computer readable program instructions expressed herein can be downloaded from the computer readable backup medium to each inference / processing grid line, or downloaded to an external computer or external backup grid line via a wireless network, such as the Internet, a local area network, a wide area network and / or a wireless network. The wireless network can include copper transmission cables, power transmission lines, wireless transmission, routers, firewalls, switches, gateway computers and / or edge service devices. The wireless network adapter card or wireless network interface in each inference / processing grid line receives the computer readable program instructions from the wireless network and forwards the computer readable program instructions for storage in the computer readable backup medium in each inference / processing grid line.

[0167] The computer program instructions for performing the operations of the present disclosure can be assembler instructions, instruction set architecture (ISA) instructions, machine instructions, machine-dependent instructions, microcode, firmware instructions, conditional setting values, or source code or object code written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Sdaletal A, H++, etc., and conventional procedural programming languages ​​such as "H" language or similar programming languages. The computer-readable program instructions can be executed entirely on the client computer, partially on the client computer, as a separate software package, partially on the client computer and partially on a remote computer, or completely on a remote computer or server. In the case of a remote computer, the remote computer can be connected to the client computer via any other wireless network, including a local area network (LAb) or a wide area network (WAb), or can be connected to an external computer (just like using an Internet service provider to connect via the Internet). In some embodiments, the electronic circuit is customized by using the state values ​​of computer-readable program instructions, such as a programmable logic circuit, a field programmable gate array (FPGA) or a programmable logic array (PLA), and the electronic circuit can execute computer-readable program instructions to achieve various aspects of the overhead disclosure.

[0168] Finally, it should be understood that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit the present invention. Although the present invention has been implemented in detail with reference to the above embodiments, ordinary technicians in the relevant field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents. Any modification or equivalent replacement that does not depart from the spirit and scope of the present invention should be included in the scope of protection of the claims of the present invention.

Claims

1. A method for identifying electricity theft based on distribution network state estimation, characterized in that: include: Step 1: Obtain measurement data from the D5000 system, power acquisition terminal messages, and power consumption information collection system; Step 2: Perform state estimation calculation on the line based on the calculated measurement values ​​at each line node; Step 3: First, arrange the measurement items, use the weight reduction method to reduce or set the weight of a measurement data to be detected, and use the reduced-weight measurement data to recalculate the line operation status. Repeat this process until all measurements are reduced-weighted and recalculated. For each reduced-weight result, obtain the residual and calculate its correction rate to determine whether there is abnormal power consumption at the measurement value node.

2. The method for identifying electricity theft based on distribution network state estimation according to claim 1, characterized in that: If the measured data in step 1 lacks relevant values ​​or contains irregular data, an interpolation algorithm is used to supplement the missing relevant values ​​or a transformation method is used to transform the irregular data into regular data.

3. The method for identifying electricity theft based on distribution network state estimation according to claim 1, characterized in that: Step 1 specifically includes: For line junctions, there is phase voltage telemetry data Specifically including A phase voltage telemetry data Phase B voltage telemetry data Phase C voltage telemetry data It can be obtained from the D5000 system; the high-voltage power meter at the gateway measures the line voltage between phase A and phase B Line voltage between phase C and phase B Line voltage between phase A and phase C Phase A current Phase B current Phase C current The total active power P, total reactive power Q and voltage phase angle information of each phase, the data measured by the high-power supply and high-power meter at the gateway are stored in the remote energy collection terminal in the station as the terminal message of the electric energy collection terminal, and can be obtained through the master station call measurement connected to the remote energy collection terminal in the station. The measurement data includes the phase voltage telemetry data of the line gateway Line voltage between phase A and phase B Line voltage between phase C and phase B Line voltage between phase A and phase C Phase A current Phase B current Phase C current Total active power P, total reactive power Q and voltage phase angle information of each phase.

4. The method for identifying electricity theft based on distribution network state estimation according to claim 3 is characterized in that: Step 1 specifically includes: for high-supply and high-meter users, the electric energy meter used as the electricity consumption information collection system measures the line voltage Phase current And the total active power P i , total reactive power Q i ; Then supplement the AC phase line voltage A, B, C phase voltage Phase B current And related phase angle information, the supplementary methods include: The three-phase voltage phase angle calculated at the line junction is used To compensate for the unknown load three-phase voltage phase angle, the voltage calculated at the line junction is used. Compensating for unknown loads Then, according to the three-variable quadratic equation system shown in formula (8): Available The measurement data also includes the line voltage for high-power and high-meter users Phase current And the total active power P i , total reactive power Q i , AC phase voltage A, B, C phase voltage Phase B current And the related voltage phase angle information.

5. The method for identifying electricity theft based on distribution network state estimation according to claim 3 is characterized in that: Step 1 specifically also includes: for the transformer branch with high supply and low metering, and the nameplate data that has been converted to the high voltage side, an equivalent mathematical model is established, and the taps of the equivalent mathematical model are all put into the corresponding gears. After the measured data is transformed, the solution of branch i-j' is equivalently converted to the solution of branch ij.

6. The method for identifying electricity theft based on distribution network state estimation according to claim 3, characterized in that: Step 1 specifically includes: adding a set of pseudo-measured values ​​of iron loss as node injection power on one side of the ideal transformer, as shown in formula (9): In the formula, ΔS T0 is the rated loss of the transformer at no load; U av is the average rated voltage of the phase; U tap is the transformer tap voltage, ΔA0 is the power injected into the node on one side of the ideal transformer, and the measured data also includes the pseudo-measurement value of iron loss.

7. The method for identifying electricity theft based on distribution network state estimation according to claim 3, characterized in that: In step 1, the measurement data also includes virtual measurement values; The measurement data also includes core volume measurements.

8. The method for identifying electricity theft based on distribution network state estimation according to claim 7, characterized in that: Step 2 specifically includes: The relationship between the measurement data and the state variables is expressed by the following measurement equation, namely formula (10): z=h(x)+e (10) Where z∈R m is the measurement vector; e=[e1,e2,……,e m ] T ∈R m is the measurement error vector; x is the state variable, which is a set of independent, minimal variables that can characterize all the characteristics of the power system. Let x = [θ T ,v T ] includes all node voltage magnitudes and phase angles (except the reference node phase angle), θ∈R N-1 is the phase angle of the node except the reference node, v∈R N is the voltage modulus of all nodes. After determining the voltage modulus and phase angle, the operating state of the entire system can be determined accordingly; h:R n →R m is the nonlinear mapping from state variables to measurement vectors, i.e., the measurement expression; The voltage measurement expression is shown in formula (11): v i =v i (11) The node injection power measurement expression is shown in formula (12): In formula (11) and formula (12), v i and θ i are the voltage magnitude and phase angle of node i respectively; θ ij =θ i -θ j ; The residual vector is expressed as shown in formula (13): In the formula, is the estimated value of the state variable vector; Based on formula (10) and the least squares method, the following mathematical model is established and solved, namely formula (14): The partial derivative of the objective function J(x) of formula (14) is calculated and set to 0, as shown in formula (15): In the formula, It is called the Jacobian matrix; Construct H(x) using formula (16): Where 0 represents a vector or matrix with all elements equal to 0; θ and v represent the phase angle vector and the voltage magnitude vector of all nodes except the reference node, respectively; v m represents the node voltage modulus measurement vector; P inj and Q inj They represent the injected active power measurement and injected reactive power measurement of the node respectively.

9. The method for identifying electricity theft based on distribution network state estimation according to claim 8, characterized in that: The specific expression of H(x) is as follows: Use formula (17) to obtain the partial derivative of the node voltage measurement for the phase angle and voltage of the node: Use formula (18) to obtain the partial derivative of the node voltage measurement with respect to the phase angle and voltage of the adjacent node: Use formula (19) to measure the active power injection of the node and obtain the partial derivative of the phase angle and voltage of the node: The active power injection measurement of the node is used to obtain the partial derivative of the phase angle and voltage of the adjacent node: The reactive power injection measurement of the node is used to obtain the partial derivative of the phase angle and voltage of the node: The reactive power injection measurement of the node is used to obtain the partial derivative of the phase angle and voltage of the adjacent node: Let the equivalent first-order optimal condition satisfied by formula (14) be g(x), which is shown in formula (23): g(x)=-H T (x)[z-h(x)] (23) Put g(x) at x (k) Perform an approximate Taylor series expansion at to linearize it, and we can get formula (24): Let g′(x)=G(x), and we get the linear form of the canonical equation, which is shown in formula (25): G(x (k) )Δx (k) =H T (x (k) )[z-h(x (k) )] (25) Where k is the number of iterations; Δx (k) =x (k+1) -x (k) ; G=H T H is the gain matrix, which is a symmetric sparse positive definite matrix; H T (x (k) )[zh(x (k) )] is the free vector; Modify the state variable x (2) =x (1) +Δx (1) , get the state variable x after the first iteration correction (2) , continue to iterate, that is, perform iterative correction according to formula (25) until the objective function J(x (k) ) until the convergence criterion is reached, which is shown in formula (26): ||Δx (k) ||≤ε (26) In the formula, ||·|| represents the infinite norm; ε is the convergence criterion, at which point the estimated values ​​of all nodes can be obtained 10. A device for identifying electricity theft based on distribution network state estimation, characterized in that: include: An acquisition module is used to obtain measurement data from the D5000 system, power acquisition terminal messages, and power consumption information acquisition system; An estimation module, which is used to perform state estimation calculation on the line according to the calculated measurement values ​​at each line node; The judgment module is used to arrange the measurement items, use the weight reduction method to reduce or set the weight of a measurement data to be detected, and recalculate the line operation status using the weighted measurement data. This process is repeated until all measurement items are weighted down and recalculated. The residual is obtained for each weight reduction result and its correction rate is calculated to determine whether there is abnormal power consumption at the measurement value node.

Images