CN104330678A - Station area electricity utilization monitoring method based on three phase state estimation of low-voltage network - Google Patents

Abstract

The invention discloses a station area electricity utilization monitoring method based on three phase state estimation of a low-voltage network. The method includes: firstly inputting three-phase voltage, three-phase current, active power data and reactive power data collected by a load terminal at a same time section and station area structural data of the low-voltage network, and initializing state estimation parameter information; then performing three phase state estimation of the low-voltage network; estimating each phase active power and reactive power of the station area load user terminals of the low-voltage network; calculating remote sensing estimation errors of each phase active power and reactive power of the station area load user terminals of the low-voltage network; comparing remote sensing estimation errors of each phase active power and reactive power of each phase power consumer of the load terminal with state estimation pass percent index values to judge whether a suspicious electricity stealing user exists.

Description

A kind of platform district electricity consumption monitoring method estimated based on low-voltage network three-phase state

Technical field

The invention belongs to distribution system electricity consumption monitoring and analysis technical field, be specifically related to a kind of platform district electricity consumption monitoring method estimated based on low-voltage network three-phase state.

Background technology

Along with the development of power industry and the construction of electrical network, Operation of Electric Systems supervisory and control means also become further flexible and ripe.But the management line loss of China's low-voltage distribution network platform region still remains high for a long time at present, brings huge electric flux loss and economic loss to power supply enterprise, wherein stealing or exception electrical phenomena are the one of the main reasons causing management line loss exception.Due to the variation of stealing mode and the reason such as Prevention Stealing Electricity Technology level is low, electrical leakage problems annoyings power supply department at different levels always steathily.Stealing or abnormal electricity consumption behavior not only very disruptive, for electricity consumption order, also have a strong impact on power supply reliability and the quality of power supply of low-voltage distribution network platform region.Therefore, in order to the just rights and interests of the economic benefit and power consumer that ensure power supply enterprise, further investigation low-voltage distribution network platform region electricity consumption monitoring method, so that Timeliness coverage and effectively containment stealing or abnormal electricity consumption user, has important researching value and realistic meaning.

For a long time, the electricity consumption monitoring of China mainly relies on manpower to check, and anti-opposing electricity-stealing mainly is designed based on " anti-electricity-theft " of user's electric energy meter, such as, adopt metering box special or metering dedicated lock, anti-picking lead sealing and lead-in wire sleeve pipe are installed additional to ammeter, adopts inverse-stopping type ammeter.The major defect of manpower on-site examination is inefficiency, and affects greatly by subjective factor, is difficult to accurate, Timeliness coverage exception electrical phenomena; In addition, " anti-electricity-theft " of electric energy meter design is mainly applicable to common stealing gimmick, as the stealing of decompression method and the stealing of undercurrent method, and wiring or the hidden electricity filching behavior such as to sunken cord before table cannot be taken precautions against.

In recent years, along with the development and apply of remote loads centralized meter-reading system, there is the electricity consumption monitoring method counting analysis based on table.As Chinese patent " a kind of electricity consumption monitoring method " (patent No. CN101477163), disclosed method first gathers voltage, electric current and power freezing data by power consumption monitoring terminal to be sent to main website, compare in main website management of computing line loss per unit and with setting threshold values and determine whether exception electrical phenomena, if discovery multiplexing electric abnormality, then utilizing nonlinear programming approach to calculate each user power, locating suspicious stealing user by comparing this rated output with the size freezing power.The major defect of the method is: 1. require that main website is according to the correct management of computing line loss per unit of image data, if line loss per unit miscount, have a strong impact on electricity consumption monitoring result, engineering practicability is poor; 2. compare size by line loss per unit with experience threshold values, determine whether exception electrical phenomena, artificial subjectivity is too strong, and the line loss per unit experience threshold values in district is general not different yet on the same stage; 3. nonlinear programming approach can not avoid the harmful effect of bad metric data, causes the result of calculation mistake of user power, cannot prepare to locate suspicious stealing user.

Summary of the invention

The object of the invention is the deficiency for existing electricity consumption monitoring method, a kind of platform district electricity consumption monitoring method estimated based on distribution three-phase state is provided.The voltage that this method gathers according to loaded termination, electric current and power data, estimated by distribution three-phase state, utilization state estimates that qualification rate index effectively identification can locate suspicious stealing user, have that identification effect is good, engineering practicability by force, be easy to the feature applied.

In order to solve above-mentioned technical matters, the invention provides a kind of platform district electricity consumption monitoring method estimated based on low-voltage network three-phase state:

Step one: three-phase voltage, three-phase current, active power data and reactive power data that input loaded termination gathers under section at one time and low-voltage distribution network platform region structured data, and init state estimated parameter information;

Step 2: carry out the estimation of low-voltage network three-phase state;

Step 3: each phase active power and the reactive power of carrying out low-voltage distribution network platform region load subscriber endpoints are estimated;

Step 4: each phase active power remote measurement evaluated error and the reactive power remote measurement evaluated error that calculate low-voltage distribution network platform region load subscriber endpoints;

Step 5: finally by comparing the active power of each phase power consumer of load end points and reactive power remote measurement evaluated error and state estimation qualification rate desired value size, determine whether suspicious stealing user.

In a preferred embodiment: described three-phase voltage, three-phase current and active power data and reactive power data comprise total active power under low-voltage distribution network platform region common transformer discontinuity surface in a period of time in office and total reactive power data and three-phase voltage amplitude and three-phase current amplitude, and each phase active power of low-voltage distribution network platform region all loads user terminal at one time under section and each phase reactive power data, three-phase voltage amplitude and phase current magnitude.

In a preferred embodiment: described low-voltage distribution network platform region structured data comprises: network line resistance, reactance and admittance parameter, the rated voltage of circuit, reference power.

In a preferred embodiment: described init state estimated parameter information specifically refers to: the three-phase node voltage amplitude arranging all end points of low-voltage distribution network platform region is 1.0pu, and a phase voltage phase place is zero, b, c two-phase voltage-phase compared with a phase voltage delayed 120 ° and 240 ° respectively; Set condition is estimated to measure weight matrix R ^-1for n rank unit matrix, i.e. n rank unit matrix R ^-1diagonal entry be 1 entirely, off diagonal element to be 0, n be entirely measurement number actual in state estimation; Initialization maximum iteration time Tmax is 40 ~ 60; Convergence precision ε is 10 ^-3~ 10 ^-5, and iterations time=1 is set.

In a preferred embodiment: described low-voltage network three-phase state is estimated specifically to comprise the following steps:

1) computational grid bus admittance matrix;

According to described electric network composition parameter, computational grid bus admittance matrix, computing formula is:

In formula: B ₁=a, b, c, n}, represent that end points comprises the electrical node combination of neutral point n, wherein a, b, c represent abc three-phase electrical node respectively; X, p, t represent electrical node combination B ₁in any node; φ _ifor being directly connected with end points i but not comprising the Extreme points set of end points i; J is Extreme points set φ _iin arbitrary end points; represent the parallel branch admittance between end points i interior joint p and x; represent the branch admittance that end points i interior joint p and end points j interior joint t is directly connected; represent the shunt admittance over the ground that end points i interior joint p ground capacitance obtains; Self-admittance represent all branch admittance and ground capacitance admittance that are directly connected with end points i interior joint p sum, transadmittance represent branch admittance between end points i interior joint p and t opposite number and end points i and and i to be directly connected all branch admittance sums between endpoint node p and t, transadmittance then represent branch admittance between end points i and k interior joint p and t opposite number.

problem 3: the letter character form of full patent texts is Times New Roman the need of unification, identical with the symbol correspondence in formula, read better to recognize? whether does (this is unwanted) alphabetical italic also need unified? if do not needed, that does not just do requirement.

2) each phase node Injection Current amplitude amount of unbalance of low-voltage distribution network platform region common transformer end points s and total active power amount of unbalance and total reactive power amount of unbalance is calculated; Computing formula is:

In formula: B _p=a, b, c}, represent that end points does not comprise the electrical node combination of neutral point n, wherein a, b, c represent abc three-phase electrical node respectively; D represents electrical node combination B _pin any node; B ₁, t the same formula of meaning (1); S represents low-voltage distribution network platform region common transformer end points; be comprise low-voltage distribution network platform region common transformer end points s and with its Extreme points set be directly connected; K is Extreme points set in arbitrary end points; be respectively and calculate the d phase node Injection Current amount of unbalance of low-voltage distribution network platform region common transformer end points s, total active power amount of unbalance and total reactive power amount of unbalance based on formula (2); with be respectively the measuring amount of low-voltage distribution network platform region common transformer end points s relative to the three-phase current amplitude of neutral point n, the total active power of three-phase and the total reactive power of three-phase; with be respectively the voltage of end points k interior joint t real part and imaginary part; with be respectively the voltage of end points s interior joint d real part and imaginary part; with be respectively the voltage of the neutral point n of end points s real part and imaginary part; with be respectively bus admittance matrix element real part and imaginary part.

3) calculate active power amount of unbalance and the reactive power amount of unbalance of the three-phase node of low-voltage distribution network platform region load subscriber endpoints i, computing formula is:

In formula: with be respectively d phase node active power amount of unbalance and reactive power amount of unbalance in low-voltage network subscriber endpoints i; with be respectively the active power of the relative neutral point n of low-voltage distribution network platform region load subscriber endpoints i interior joint d and the transient measurement amount of reactive power.I represents certain load subscriber endpoints any in low-voltage distribution network platform region; with be respectively the voltage of end points i interior joint d real part and imaginary part; with be respectively the neutral point n voltage of end points i real part and imaginary part; with be respectively bus admittance matrix element real part and imaginary part; be comprise load subscriber endpoints i and with its set of end points that is directly connected; K is Extreme points set in arbitrary end points; b _p, d, B ₁and the same formula of the meaning of t (2).

4) calculate active power amount of unbalance and the reactive power amount of unbalance of the neutral point of low-voltage distribution network platform region load subscriber endpoints i, computing formula is:

5) calculate the amount of unbalance of the three-phase node voltage of low-voltage distribution network platform region load subscriber endpoints i, computing formula is:

$\mathrm{\Δ}{z}_{i,U}^d={\left({\hat{U}}_i^{\mathrm{dn}}\right)}^2-({(e_i^d-e_i^n)}^2+{({f_i}^d-{f_i}^n)}^2),d\∈B_P---\left(5\right)$

4) and 5) in formula: with be respectively the active power amount of unbalance of the neutral point n of low-voltage distribution network platform region load subscriber endpoints i and the amount of unbalance of reactive power; for the Voltage unbalance amount of d phase node in low-voltage distribution network load subscriber endpoints i; for the measuring amount of the voltage magnitude of the relative neutral point n of low-voltage distribution network platform region load subscriber endpoints i; with be respectively bus admittance matrix element real part and imaginary part; k, B _p, d, B ₁and the same formula of the meaning of t (3).

6) the current imbalance amount of three-phase node and the current imbalance amount Δ c of neutral point in contact zero end points m is calculated _m, its computing formula is:

In formula: with be respectively bus admittance matrix element real part and imaginary part; be comprise contact zero end points m and with its set of end points that is directly connected; K is Extreme points set in arbitrary end points; the same formula of meaning (2); B ₁, t, p the same formula of meaning (1).

7) measurement Jacobian matrix is calculated;

Form measurement jacobian matrix H=[H _ss, H _sj; H _is, H _ij], the computing formula of each piecemeal submatrix of Jacobian matrix H is:

In formula: Jacobi's piecemeal submatrix H _ss=[H _ss1(1:3,1:3); H _ss2(1:3,1:3); H _ss3(1,1:3); H _ss4(1,1:3)]; with be respectively bus admittance matrix element with real part, with be respectively bus admittance matrix element with imaginary part; with be respectively end points f interior joint t voltage real part and imaginary part, f is Extreme points set in arbitrary end points; R represents electrical node combination B _pin any node; with be respectively the node r voltage in end points s real part and imaginary part; b _p, B ₁, d, t, s the same formula of meaning (4).

$\left\{\begin{array}{c}{H}_{\mathrm{sj}1}(1:\mathrm{3,1}:8)={\left[\begin{array}{c}2G_{\mathrm{sj}}^{\mathrm{dt}}(G_{\mathrm{sj}}^{\mathrm{dt}}{e}_j^t-B_{\mathrm{sj}}^{\mathrm{dt}}{f}_j^t)+2B_{\mathrm{sj}}^{\mathrm{dt}}(G_{\mathrm{sj}}^{\mathrm{dt}}{f}_j^t+B_{\mathrm{sj}}^{\mathrm{dt}}{e}_j^t)\\ -2B_{\mathrm{sj}}^{\mathrm{dt}}(G_{\mathrm{sj}}^{\mathrm{dt}}{e}_j^t-B_{\mathrm{sj}}^{\mathrm{dt}}{f}_j^t)+2G_{\mathrm{sj}}^{\mathrm{dt}}(G_{\mathrm{sj}}^{\mathrm{dt}}{f}_j^t+B_{\mathrm{sj}}^{\mathrm{dt}}{e}_j^t)\end{array}\right]}_{3\×8}^T\\ H_{\mathrm{sj}2}(1:\mathrm{3,1}:8)={\left[0\right]}_{3\×8}\\ H_{\mathrm{sj}3}(\mathrm{1,1}:8)={\left[\begin{array}{c}\underset{d\∈B_P}{\mathrm{\Σ}}\{G_{\mathrm{sj}}^{\mathrm{dt}}(e_s^d-e_s^n)+B_{\mathrm{sj}}^{\mathrm{dt}}({f_s}^d-{f_s}^n)\}\\ \underset{d\∈B_P}{\mathrm{\Σ}}\{-B_{\mathrm{sj}}^{\mathrm{dt}}(e_s^d-e_s^n)+G_{\mathrm{sj}}^{\mathrm{dt}}({f_s}^d-{f_s}^n)\}\end{array}\right]}_{1\×8}^T\\ H_{\mathrm{sj}4}(\mathrm{1,1}:8)={\left[\begin{array}{c}\underset{d\∈B_P}{\mathrm{\Σ}}\{-B_{\mathrm{sj}}^{\mathrm{dt}}(e_s^d-e_s^n)+G_{\mathrm{sj}}^{\mathrm{dt}}({f_s}^d-{f_s}^n)\}\\ \underset{d\∈B_P}{\mathrm{\Σ}}\{-G_{\mathrm{sj}}^{\mathrm{dt}}(e_s^d-e_s^n)-B_{\mathrm{sj}}^{\mathrm{dt}}({f_s}^d-{f_s}^n)\}\end{array}\right]}_{1\×8}^T\end{array}\right.---\left(8\right)$

In formula: Jacobi's piecemeal submatrix H _sj=[H _sj1(1:3,1:8); H _sj2(1:3,1:8); H _sj3(1,1:8); H _sj4(1,1:8)]; S is low-voltage distribution network platform region common transformer end points; J is other end points except low-voltage distribution network platform region common transformer end points s; with be respectively bus admittance matrix element real part and imaginary part; with be respectively end points j interior joint t voltage real part and imaginary part; B _p, the same formula of meaning (2) of d and t; [0] _{3 × 8}represent full null matrix, subscript 3 × 8 represents that this matrix is that 3 row 8 arrange.

$\left\{\begin{array}{c}{H}_{\mathrm{is}}(1:\mathrm{8,1}:6)={\left[\begin{array}{cc}{G}_{\mathrm{is}}^{\mathrm{pd}}& -B_{\mathrm{is}}^{\mathrm{pd}}\\ B_{\mathrm{is}}^{\mathrm{pd}}& G_{\mathrm{is}}^{\mathrm{pd}}\end{array}\right]}_{8\×6},p\∈B_1,d\∈B_P\\ H_{\mathrm{is}}(9:\mathrm{11,1}:6)={\left[\begin{array}{c}0\end{array}\right]}_{3\×6}\end{array}\right.---\left(9\right)$

In formula: with be respectively bus admittance matrix element real part and imaginary part; S is low-voltage distribution network platform region common transformer end points; I is low-voltage distribution network platform region load subscriber endpoints; [0] _{3 × 6}represent full null matrix, subscript 3 × 6 represents that this matrix is that 3 row 6 arrange; P, B _p, B ₁the same formula of meaning (1); The same formula of meaning (2) of d.

$\left\{\begin{array}{c}{H}_{\mathrm{ij}}(1:\mathrm{8,1}:8)={\left[\begin{array}{cc}{G}_{\mathrm{ij}}^{\mathrm{pt}}& -B_{\mathrm{ij}}^{\mathrm{pt}}\\ B_{\mathrm{ij}}^{\mathrm{pt}}& G_{\mathrm{ij}}^{\mathrm{pt}}\end{array}\right]}_{8\×8},p\∈B_1,t\∈B_1\\ H_{\mathrm{ij}}(9:\mathrm{11,1}:8)={\left[\begin{array}{c}0\end{array}\right]}_{3\×8}\end{array}\right.---\left(10\right)$

In formula: H _ijfor the constant term piecemeal submatrix of Jacobi matrix H; with be respectively bus admittance matrix element real part and imaginary part; [0] _{3 × 8}the same formula of meaning (8); P, t and B ₁the same formula of meaning (1).

$\left\{\begin{array}{c}\mathrm{\Δ}{H}_{\mathrm{ii}}(1:\mathrm{3,1}:3)=-\mathrm{\Δ}{H}_{\mathrm{ii}}(5:\mathrm{7,5}:7)={\left[{\mathrm{\μ}}_i^d\right]}_{3\×3}\\ \mathrm{\Δ}{H}_{\mathrm{ii}}(5:\mathrm{7,8})=-\mathrm{\Δ}{H}_{\mathrm{ii}}(1:\mathrm{3,4})={\left[{\mathrm{\μ}}_i^d\right]}_{3\×1}\\ \mathrm{\Δ}{H}_{\mathrm{ii}}(\mathrm{8,5}:7)=-\mathrm{\Δ}{H}_{\mathrm{ii}}(\mathrm{4,1}:3)={\left[{\mathrm{\μ}}_i^d\right]}_{1\×3}\\ \mathrm{\Δ}{H}_{\mathrm{ii}}\left(\mathrm{4,4}\right)=-\mathrm{\Δ}{H}_{\mathrm{ii}}\left(\mathrm{8,8}\right)=\underset{d\∈B_P}{\mathrm{\Σ}}{\mathrm{\μ}}_i^d\\ \mathrm{\Δ}{H}_{\mathrm{ii}}(1:\mathrm{3,5}:7)=\mathrm{\Δ}{H}_{\mathrm{ii}}(5:\mathrm{7,1}:3)={[-{\mathrm{\η}}_i^d]}_{3\×3}\\ \mathrm{\Δ}{H}_{\mathrm{ii}}(1:\mathrm{3,8})=\mathrm{\Δ}{H}_{\mathrm{ii}}(5:\mathrm{7,4})={\left[{\mathrm{\η}}_i^d\right]}_{3\×1}\\ \mathrm{\Δ}{H}_{\mathrm{ii}}(\mathrm{4,5}:7)=\mathrm{\Δ}{H}_{\mathrm{ii}}(\mathrm{8,1}:3)={\left[{\mathrm{\η}}_i^d\right]}_{1\×3}\\ \mathrm{\Δ}{H}_{\mathrm{ii}}\left(\mathrm{4,8}\right)=\mathrm{\Δ}{H}_{\mathrm{ii}}\left(\mathrm{8,4}\right)=-\underset{d\∈B_P}{\mathrm{\Σ}}{\mathrm{\η}}_i^d\\ \mathrm{\Δ}{H}_{\mathrm{ii}}(9:\mathrm{11,1}:3)={\left[2(e_i^d-e_i^n)\right]}_{3\×3}\\ \mathrm{\Δ}{H}_{\mathrm{ii}}(9:\mathrm{11,4})={[-2(e_i^d-e_i^n)]}_{3\×1}\\ \mathrm{\Δ}{H}_{\mathrm{ii}}(9:\mathrm{11,5}:7)={\left[2({f_i}^d-{f_i}^n)\right]}_{3\×3}\\ \mathrm{\Δ}{H}_{\mathrm{ii}}(9:\mathrm{11,8})={[-2({f_i}^d-{f_i}^n)]}_{3\×1}\end{array}\right.---\left(11\right)$

In formula: Δ H _iifor Jacobi's piecemeal submatrix H _iicorrection term submatrix, i is other end points except low-voltage distribution network platform region common transformer; B _p, d, and the same formula of meaning (3); with be correction term submatrix Δ H _iimatrix element, its computing formula is respectively:

$\left\{\begin{array}{c}{\mathrm{\η}}_i^d=\frac{{\hat{Q}}_i^{\mathrm{dn}}({(e_i^d-e_i^n)}^2-{({f_i}^d-{f_i}^n)}^2)-2{\hat{P}}_i^{\mathrm{dn}}(e_i^d-e_i^n)({f_i}^d-{f_i}^n)}{{({(e_i^d-e_i^n)}^2+{({f_i}^d-{f_i}^n)}^2)}^2}\\ {\mathrm{\μ}}_i^d=\frac{{\hat{P}}_i^{\mathrm{dn}}({(e_i^d-e_i^n)}^2-{({f_i}^d-{f_i}^n)}^2)+2{\hat{Q}}_i^{\mathrm{dn}}(e_i^d-e_i^n)({f_i}^d-{f_i}^n)}{{({(e_i^d-e_i^n)}^2+{({f_i}^d-{f_i}^n)}^2)}^2}\end{array}\right.,d\∈B_P---\left(12\right)$

In formula: B _p, d, and the same formula of meaning (3).

Calculate Jacobi matrix C, computing formula is:

$C_{\mathrm{ij}}(1:\mathrm{8,1}:8)={\left[\begin{array}{cc}{G}_{\mathrm{ij}}^{\mathrm{dt}}& -B_{\mathrm{ij}}^{\mathrm{dt}}\\ B_{\mathrm{ij}}^{\mathrm{dt}}& G_{\mathrm{ij}}^{\mathrm{dt}}\end{array}\right]}_{8\×8},d\∈B_1,t\∈B_1---\left(13\right)$

In formula: C _ijfor the submatrix of Jacobi matrix C; I, j, d, t and B ₁the same formula of meaning (10).

8) state variable upgrades: the correction amount x of computing mode variable ^(time), then upgrade state variable, obtain state variable and be newly worth; That is: x ^(time+1)=x ^(time)+ Δ x ^(time), time=time+1, correction computing formula is:

Δ x=(H ^tr ^-1h) ^-1{ H ^tr ^-1Δ z-C ^t[C (H ^tr ^-1h) ^-1c ^t] ^-1[C (H ^tr ^-1h) ^-1h ^tr ^-1Δ z-Δ c] } in (14) formula, time is for calculating iterations; H ^tand C ^tbe respectively the transposition of Jacobian matrix H and C; Δ c=-c (x); represent the measurement amount of unbalance that iterative value calculates based on formula (2)-(5) when being x, comprise injection that low-voltage distribution network platform region public affairs become three-phase node and neutral point in end points s each phase node Injection Current amplitude amount of unbalance and total active power amount of unbalance and total reactive power amount of unbalance, low-voltage network load subscriber endpoints i and to gain merit unbalanced power amount and inject amount of unbalance and the Voltage unbalance amount of reactive power; R ^-1for n rank unit matrix, the same formula of its meaning (1).

9) state estimation convergence judges: if the correction amount x of state variable ^(time)meet max (| Δ x ^(time)|) < ε, then finishing iteration calculates, and goes to step three; If max (| Δ x ^(time)|)>=ε and iterations time≤Tmax, then return step 2, continue iteration and carry out state estimation; When max (| Δ x ^(time)|)>=ε and iterations time>Tmax, then directly stop iteration, export " state estimation does not restrain ".

In a preferred embodiment: the described active power of low-voltage distribution network platform region load subscriber endpoints and the computing formula of reactive power estimated value are:

In formula: with be respectively the active power of d phase power consumer and the estimated value of reactive power of low-voltage distribution network platform region load subscriber endpoints i; The same formula of remaining variables meaning (3).

In a preferred embodiment: each phase active power of described low-voltage distribution network platform region load subscriber endpoints and the computing formula of reactive power remote measurement evaluated error are:

$\left\{\begin{array}{c}{\mathrm{\&lambda;}}_{i,P}^d=|P_i^{\mathrm{dn}}-{\hat{P}}_i^{\mathrm{dn}}|/Z_i\\ {\mathrm{\&lambda;}}_{i,Q}^d=|Q_i^{\mathrm{dn}}-{\hat{Q}}_i^{\mathrm{dn}}|/Z_i\end{array}\right.,d\&Element;B_P---\left(16\right)$

In formula, with the d phase active power of low-voltage distribution network platform region load subscriber endpoints i and the remote measurement evaluated error of reactive power respectively; with same formula (3); with same formula (15); Z _ifor the active power of end points i or the measurement type reference value of reactive power, be for 380/220V low-voltage platform area electrical network herein, the measurement type reference value of active power and reactive power is all 145kW.

In a preferred embodiment: the active power of described each phase power consumer by comparing load end points and reactive power remote measurement evaluated error and state estimation qualification rate desired value size, determine whether suspicious stealing user specifically to refer to: if the d phase active power of low-voltage distribution network platform region load subscriber endpoints i or the remote measurement evaluated error of reactive power meet formula (17), then it is determined that the presence of stealing or exception electrical phenomena, the d phase power consumer namely picking out low-voltage distribution network platform region load subscriber endpoints i is suspicious stealing user; Otherwise normal electricity consumption user thought by the d phase power consumer of low-voltage distribution network platform region load subscriber endpoints i.

${\mathrm{\&lambda;}}_{i,P}^d\&GreaterEqual;2.5\%,\mathrm{or}{\mathrm{\&lambda;}}_{i,Q}^d\&GreaterEqual;2.5\%---\left(17\right)$

Compared to prior art, technical scheme of the present invention possesses following beneficial effect:

1. the inventive method takes full advantage of the redundancy metric data that remote centralized meter-reading system gathers, and can effectively avoid bad metric data on the impact of suspicious stealing user discrimination result, identification result accuracy is high.

2. the inventive method is based on distribution three-phase state estimated result, and utilization state estimates that qualification rate index comes identification and the suspicious stealing user in location, and theoretical foundation is abundant, and identification effect is good.

3. the present invention can be widely used in electricity consumption monitoring and the analysis field of low-voltage distribution network platform region, locates abnormal electricity consumption behavior or stealing user provides technological means and practical approach for accurate recognition, has good practical value and application on site prospect.

Accompanying drawing explanation

Fig. 1 is for checking the schematic top plan view of the pick-up unit in main transformer protection Trip Logic loop in the preferred embodiment of the present invention;

Fig. 2 is the circuit principle structure schematic diagram that invention is preferably implemented.

Embodiment

Hereafter the present invention will be further described with specific embodiment by reference to the accompanying drawings.

As shown in Figure 1, a kind of concrete steps of the platform district electricity consumption monitoring method based on the estimation of low-voltage network three-phase state are as follows:

Step one:

1) three-phase voltage, three-phase current, active power data and reactive power data and low-voltage distribution network platform region structured data that loaded termination gathers under section are at one time inputted.

Comprise total active power under low-voltage distribution network platform region common transformer discontinuity surface in a period of time in office and total reactive power data and three-phase voltage amplitude and three-phase current amplitude, and each phase active power of low-voltage distribution network platform region all loads user terminal at one time under section and each phase reactive power data, three-phase voltage amplitude and phase current magnitude.Described low-voltage distribution network platform region structured data comprises: network line resistance, reactance and admittance parameter, the rated voltage of circuit, reference power.

2) init state estimated parameter information

The three-phase node voltage amplitude arranging all end points of low-voltage distribution network platform region is 1.0pu, and a phase voltage phase place is zero, b, c two-phase voltage-phase compared with a phase voltage delayed 120 ° and 240 ° respectively; Set condition is estimated to measure weight matrix R ^-1for n rank unit matrix, i.e. n rank unit matrix R ^-1diagonal entry be 1 entirely, off diagonal element to be 0, n be entirely measurement number actual in state estimation; Initialization maximum iteration time Tmax is 50; Convergence precision ε is 10 ^-4, and iterations time=1 is set.In setting IEEE13 node system, end points 5 is contact node in the middle of circuit, and do not access electric load user, the injection load power namely getting in touch with end points is zero.

Step 2: low-voltage network three-phase state is estimated

Concrete steps are as follows:

(1) computational grid bus admittance matrix

According to described electric network composition parameter, computational grid bus admittance matrix, computing formula is:

In formula: B ₁=a, b, c, n}, represent that end points comprises the electrical node combination of neutral point n, wherein a, b, c represent abc three-phase electrical node respectively; X, p, t represent electrical node combination B ₁in any node; φ _ifor being directly connected with end points i but not comprising the Extreme points set of end points i; J is Extreme points set φ _iin arbitrary end points; represent the parallel branch admittance between end points i interior joint p and x; represent the branch admittance that end points i interior joint p and end points j interior joint t is directly connected; represent the ground capacitance admittance that end points i interior joint p ground capacitance obtains; Self-admittance represent all branch admittance and ground capacitance admittance that are directly connected with end points i interior joint p sum, transadmittance represent branch admittance between end points i interior joint p and t opposite number and end points i and and i to be directly connected all branch admittance sums between endpoint node p and t, transadmittance then represent branch admittance between end points i and k interior joint p and t opposite number.

Calculate the bus admittance matrix Y of this network

$Y={\left[\begin{array}{ccccc}1.2123-3.7440i& -0.4466+0.8059i& ...& 0& 0\\ -0.4466+0.8059i& 1.1518-3.6853i& ...& 0& 0\\ ...& ...& ...& ...& ...\\ 0& 0& ...& 4.4304-14.5739i& -0.8396+1.9169i\\ 0& 0& ...& -0.9486+1.9540i& 4.2288-14.4533i\end{array}\right]}_{52\&times;52}$

(2) each phase node Injection Current amplitude amount of unbalance of low-voltage distribution network platform region common transformer end points s and total active power amount of unbalance and total reactive power amount of unbalance is calculated; Computing formula is:

In formula: B _p=a, b, c}, represent that end points does not comprise the electrical node combination of neutral point n, wherein a, b, c represent abc three-phase electrical node respectively; D represents electrical node combination B _pin any node; B ₁, t the same formula of meaning (1); S represents low-voltage distribution network platform region common transformer end points; be comprise low-voltage distribution network platform region common transformer end points s and with its Extreme points set be directly connected; K is Extreme points set in arbitrary end points; with be respectively and calculate the d phase node Injection Current amount of unbalance of low-voltage distribution network platform region common transformer end points s, total active power amount of unbalance and total reactive power amount of unbalance based on formula (2); with be respectively the measuring amount of low-voltage distribution network platform region common transformer end points s relative to the three-phase current amplitude of neutral point n, the total active power of three-phase and the total reactive power of three-phase; with be respectively the voltage of end points k interior joint t real part and imaginary part; with be respectively the voltage of end points s interior joint d real part and imaginary part; with be respectively the voltage of the neutral point n of end points s real part and imaginary part; with be respectively bus admittance matrix element real part and imaginary part.

(3) calculate active power amount of unbalance and the reactive power amount of unbalance of the three-phase node of low-voltage distribution network platform region load subscriber endpoints i, computing formula is:

In formula: with be respectively d phase node in low-voltage distribution network platform region load subscriber endpoints i and inject meritorious unbalanced power amount and reactive power amount of unbalance; with be respectively the active power of the relative neutral point n of low-voltage distribution network platform region load subscriber endpoints i interior joint d and the transient measurement amount of reactive power.I represents certain load subscriber endpoints any in low-voltage distribution network platform region; with be respectively the voltage of end points i interior joint d real part and imaginary part; with be respectively the neutral point n voltage of end points i real part and imaginary part; with be respectively bus admittance matrix element real part and imaginary part; be comprise load subscriber endpoints i and with its set of end points that is directly connected; K is Extreme points set in arbitrary end points; b _p, d, B ₁and the same formula of the meaning of t (2).

(4) calculate active power amount of unbalance and the reactive power amount of unbalance of the neutral point of low-voltage distribution network platform region load subscriber endpoints i, computing formula is:

(5) calculate the amount of unbalance of the three-phase node voltage of low-voltage distribution network platform region load subscriber endpoints i, computing formula is respectively:

$\mathrm{\&Delta;}{z}_{i,U}^d={\left({\hat{U}}_i^{\mathrm{dn}}\right)}^2-({(e_i^d-e_i^n)}^2+{({f_i}^d-{f_i}^n)}^2),d\&Element;B_P---\left(5\right)$

(4) and in (5) formula: with the injection being respectively the neutral point n of low-voltage distribution network platform region load subscriber endpoints i is gained merit unbalanced power amount and inject the amount of unbalance of reactive power; for the Voltage unbalance amount of d phase node in low-voltage distribution network platform district load subscriber endpoints i; for the measuring amount of the voltage magnitude of the relative neutral point n of low-voltage distribution network platform region load subscriber endpoints i; with be respectively bus admittance matrix element real part and imaginary part; k, B _p, d, B ₁and the same formula of the meaning of t (3).

(6) the current imbalance amount of three-phase node and the current imbalance amount Δ c of neutral point in contact zero end points m is calculated _m, its computing formula is:

In formula: with be respectively bus admittance matrix element real part and imaginary part; be comprise contact zero end points m and with its set of end points that is directly connected; K is Extreme points set in arbitrary end points; the same formula of meaning (2); B ₁, t, p the same formula of meaning (1).

With the citing of the result of the 1st iterative computation, by formula (2)-formula (6), calculate amount of unbalance Δ z and Δ c is respectively:

$\mathrm{\&Delta;z}={\left[\begin{array}{c}4.6426\\ 3.7631\\ -2.1184\\ 0.0168\\ .\\ .\\ .\\ 0.0017\\ \begin{array}{c}0.0021\\ 0.0015\end{array}\end{array}\right]}_{129\&times;1};$ $\mathrm{\&Delta;c}={\left[\begin{array}{c}2.5452\&times;{10}^{-5}\\ 1.0926\&times;{10}^{-7}\\ -5.4670\&times;{10}^{-5}\\ -1.2967\&times;{10}^{-5}\\ -3.5310\&times;{10}^{-5}\\ -4.2814\&times;{10}^{-5}\\ 1.3557\&times;{10}^{-4}\\ -4.6875\&times;{10}^{-5}\end{array}\right]}_{8\&times;1}$

(7) measurement Jacobian matrix is calculated

Form measurement jacobian matrix H=[H _ss, H _sj; H _is, H _ij], the computing formula of each piecemeal submatrix of Jacobian matrix H is:

In formula: Jacobi's piecemeal submatrix H _ss=[H _ss1(1:3,1:3); H _ss2(1:3,1:3); H _ss3(1,1:3); H _ss4(1,1:3)]; with be respectively bus admittance matrix element with real part, with be respectively bus admittance matrix element with imaginary part; with be respectively end points f interior joint t voltage real part and imaginary part, f is Extreme points set in arbitrary end points; R represents electrical node combination B _pin any node; with be respectively the node r voltage in end points s real part and imaginary part; b _p, B ₁, d, t, s the same formula of meaning (4).

$\left\{\begin{array}{c}{H}_{\mathrm{sj}1}(1:\mathrm{3,1}:8)={\left[\begin{array}{c}2G_{\mathrm{sj}}^{\mathrm{dt}}(G_{\mathrm{sj}}^{\mathrm{dt}}{e}_j^t-B_{\mathrm{sj}}^{\mathrm{dt}}{f}_j^t)+2B_{\mathrm{sj}}^{\mathrm{dt}}(G_{\mathrm{sj}}^{\mathrm{dt}}{f}_j^t+B_{\mathrm{sj}}^{\mathrm{dt}}{e}_j^t)\\ -2B_{\mathrm{sj}}^{\mathrm{dt}}(G_{\mathrm{sj}}^{\mathrm{dt}}{e}_j^t-B_{\mathrm{sj}}^{\mathrm{dt}}{f}_j^t)+2G_{\mathrm{sj}}^{\mathrm{dt}}(G_{\mathrm{sj}}^{\mathrm{dt}}{f}_j^t+B_{\mathrm{sj}}^{\mathrm{dt}}{e}_j^t)\end{array}\right]}_{3\&times;8}^T\\ H_{\mathrm{sj}2}(1:\mathrm{3,1}:8)={\left[0\right]}_{3\&times;8}\\ H_{\mathrm{sj}3}(\mathrm{1,1}:8)={\left[\begin{array}{c}\underset{d\&Element;B_P}{\mathrm{\&Sigma;}}\{G_{\mathrm{sj}}^{\mathrm{dt}}(e_s^d-e_s^n)+B_{\mathrm{sj}}^{\mathrm{dt}}({f_s}^d-{f_s}^n)\}\\ \underset{d\&Element;B_P}{\mathrm{\&Sigma;}}\{-B_{\mathrm{sj}}^{\mathrm{dt}}(e_s^d-e_s^n)+G_{\mathrm{sj}}^{\mathrm{dt}}({f_s}^d-{f_s}^n)\}\end{array}\right]}_{1\&times;8}^T\\ H_{\mathrm{sj}4}(\mathrm{1,1}:8)={\left[\begin{array}{c}\underset{d\&Element;B_P}{\mathrm{\&Sigma;}}\{-B_{\mathrm{sj}}^{\mathrm{dt}}(e_s^d-e_s^n)+G_{\mathrm{sj}}^{\mathrm{dt}}({f_s}^d-{f_s}^n)\}\\ \underset{d\&Element;B_P}{\mathrm{\&Sigma;}}\{-G_{\mathrm{sj}}^{\mathrm{dt}}(e_s^d-e_s^n)-B_{\mathrm{sj}}^{\mathrm{dt}}({f_s}^d-{f_s}^n)\}\end{array}\right]}_{1\&times;8}^T\end{array}\right.---\left(8\right)$

In formula: Jacobi's piecemeal submatrix H _sj=[H _sj1(1:3,1:8); H _sj2(1:3,1:8); H _sj3(1,1:8); H _sj4(1,1:8)]; S is low-voltage distribution network platform region common transformer end points; J is other end points except low-voltage distribution network platform region common transformer end points s; with be respectively bus admittance matrix element real part and imaginary part; with be respectively end points j interior joint t voltage real part and imaginary part; B _p, the same formula of meaning (2) of d and t; [0] _{3 × 8}represent full null matrix, subscript 3 × 8 represents that this matrix is that 3 row 8 arrange.

$\left\{\begin{array}{c}{H}_{\mathrm{is}}(1:\mathrm{8,1}:6)={\left[\begin{array}{cc}{G}_{\mathrm{is}}^{\mathrm{pd}}& -B_{\mathrm{is}}^{\mathrm{pd}}\\ B_{\mathrm{is}}^{\mathrm{pd}}& G_{\mathrm{is}}^{\mathrm{pd}}\end{array}\right]}_{8\&times;6},p\&Element;B_1,d\&Element;B_P\\ H_{\mathrm{is}}(9:\mathrm{11,1}:6)={\left[\begin{array}{c}0\end{array}\right]}_{3\&times;6}\end{array}\right.---\left(9\right)$

In formula: with be respectively bus admittance matrix element real part and imaginary part; S is low-voltage distribution network platform region common transformer end points; I is low-voltage distribution network platform region load subscriber endpoints; [0] _{3 × 6}represent full null matrix, subscript 3 × 6 represents that this matrix is that 3 row 6 arrange; P, B _p, B ₁the same formula of meaning (1); The same formula of meaning (2) of d.

$\left\{\begin{array}{c}{H}_{\mathrm{ij}}(1:\mathrm{8,1}:8)={\left[\begin{array}{cc}{G}_{\mathrm{ij}}^{\mathrm{pt}}& -B_{\mathrm{ij}}^{\mathrm{pt}}\\ B_{\mathrm{ij}}^{\mathrm{pt}}& G_{\mathrm{ij}}^{\mathrm{pt}}\end{array}\right]}_{8\&times;8},p\&Element;B_1,t\&Element;B_1\\ H_{\mathrm{ij}}(9:\mathrm{11,1}:8)={\left[\begin{array}{c}0\end{array}\right]}_{3\&times;8}\end{array}\right.---\left(10\right)$

In formula: H _ijfor the constant term piecemeal submatrix of Jacobi matrix H; with be respectively bus admittance matrix element real part and imaginary part; [0] _{3 × 8}the same formula of meaning (8); P, t and B ₁the same formula of meaning (1).

$\left\{\begin{array}{c}\mathrm{\&Delta;}{H}_{\mathrm{ii}}(1:\mathrm{3,1}:3)=-\mathrm{\&Delta;}{H}_{\mathrm{ii}}(5:\mathrm{7,5}:7)={\left[{\mathrm{\&mu;}}_i^d\right]}_{3\&times;3}\\ \mathrm{\&Delta;}{H}_{\mathrm{ii}}(5:\mathrm{7,8})=-\mathrm{\&Delta;}{H}_{\mathrm{ii}}(1:\mathrm{3,4})={\left[{\mathrm{\&mu;}}_i^d\right]}_{3\&times;1}\\ \mathrm{\&Delta;}{H}_{\mathrm{ii}}(\mathrm{8,5}:7)=-\mathrm{\&Delta;}{H}_{\mathrm{ii}}(\mathrm{4,1}:3)={\left[{\mathrm{\&mu;}}_i^d\right]}_{1\&times;3}\\ \mathrm{\&Delta;}{H}_{\mathrm{ii}}\left(\mathrm{4,4}\right)=-\mathrm{\&Delta;}{H}_{\mathrm{ii}}\left(\mathrm{8,8}\right)=\underset{d\&Element;B_P}{\mathrm{\&Sigma;}}{\mathrm{\&mu;}}_i^d\\ \mathrm{\&Delta;}{H}_{\mathrm{ii}}(1:\mathrm{3,5}:7)=\mathrm{\&Delta;}{H}_{\mathrm{ii}}(5:\mathrm{7,1}:3)={[-{\mathrm{\&eta;}}_i^d]}_{3\&times;3}\\ \mathrm{\&Delta;}{H}_{\mathrm{ii}}(1:\mathrm{3,8})=\mathrm{\&Delta;}{H}_{\mathrm{ii}}(5:\mathrm{7,4})={\left[{\mathrm{\&eta;}}_i^d\right]}_{3\&times;1}\\ \mathrm{\&Delta;}{H}_{\mathrm{ii}}(\mathrm{4,5}:7)=\mathrm{\&Delta;}{H}_{\mathrm{ii}}(\mathrm{8,1}:3)={\left[{\mathrm{\&eta;}}_i^d\right]}_{1\&times;3}\\ \mathrm{\&Delta;}{H}_{\mathrm{ii}}\left(\mathrm{4,8}\right)=\mathrm{\&Delta;}{H}_{\mathrm{ii}}\left(\mathrm{8,4}\right)=-\underset{d\&Element;B_P}{\mathrm{\&Sigma;}}{\mathrm{\&eta;}}_i^d\\ \mathrm{\&Delta;}{H}_{\mathrm{ii}}(9:\mathrm{11,1}:3)={\left[2(e_i^d-e_i^n)\right]}_{3\&times;3}\\ \mathrm{\&Delta;}{H}_{\mathrm{ii}}(9:\mathrm{11,4})={[-2(e_i^d-e_i^n)]}_{3\&times;1}\\ \mathrm{\&Delta;}{H}_{\mathrm{ii}}(9:\mathrm{11,5}:7)={\left[2({f_i}^d-{f_i}^n)\right]}_{3\&times;3}\\ \mathrm{\&Delta;}{H}_{\mathrm{ii}}(9:\mathrm{11,8})={[-2({f_i}^d-{f_i}^n)]}_{3\&times;1}\end{array}\right.---\left(11\right)$

In formula: Δ H _iifor Jacobi's piecemeal submatrix H _iicorrection term submatrix, i is other end points except low-voltage distribution network platform region common transformer; B _p, d, and the same formula of meaning (3); with be correction term submatrix Δ H _iimatrix element, its computing formula is respectively:

$\left\{\begin{array}{c}{\mathrm{\&eta;}}_i^d=\frac{{\hat{Q}}_i^{\mathrm{dn}}({(e_i^d-e_i^n)}^2-{({f_i}^d-{f_i}^n)}^2)-2{\hat{P}}_i^{\mathrm{dn}}(e_i^d-e_i^n)({f_i}^d-{f_i}^n)}{{({(e_i^d-e_i^n)}^2+{({f_i}^d-{f_i}^n)}^2)}^2}\\ {\mathrm{\&mu;}}_i^d=\frac{{\hat{P}}_i^{\mathrm{dn}}({(e_i^d-e_i^n)}^2-{({f_i}^d-{f_i}^n)}^2)+2{\hat{Q}}_i^{\mathrm{dn}}(e_i^d-e_i^n)({f_i}^d-{f_i}^n)}{{({(e_i^d-e_i^n)}^2+{({f_i}^d-{f_i}^n)}^2)}^2}\end{array}\right.,d\&Element;B_P---\left(12\right)$

In formula: B _p, d, and the same formula of meaning (3).

Calculate Jacobi matrix C, computing formula is:

$C_{\mathrm{ij}}(1:\mathrm{8,1}:8)={\left[\begin{array}{cc}{G}_{\mathrm{ij}}^{\mathrm{dt}}& -B_{\mathrm{ij}}^{\mathrm{dt}}\\ B_{\mathrm{ij}}^{\mathrm{dt}}& G_{\mathrm{ij}}^{\mathrm{dt}}\end{array}\right]}_{8\&times;8},d\&Element;B_1,t\&Element;B_1---\left(13\right)$

In formula: C _ijfor the submatrix of Jacobi matrix C; I, j, d, t and B ₁the same formula of meaning (10).

With the citing of the result of the 1st iterative computation, by formula (7)-formula (12), calculating measurement jacobian matrix H is:

$H={\left[\begin{array}{ccccccc}7.439\&times;{10}^4& -2.001\&times;{10}^3& -1.481\&times;{10}^3& ...& 0& 0& 0\\ 4.077\&times;{10}^3& -3.514\&times;{10}^4& -0.801\&times;{10}^3& ...& 0& 0& 0\\ 2.869\&times;{10}^3& -0.762\&times;{10}^3& -3.595\&times;{10}^4& ...& 0& 0& 0\\ .& .& .& .& .& .& .\\ .& .& .& .& .& .& .\\ .& .& .& .& .& .& .\\ 0& 0& 0& ...& 0& 0& 531.041\\ 0& 0& 0& ..& -4.161\&times;{10}^3& 0& 4.161\&times;{10}^3\\ 0& 0& 0& ...& 0& 3.598\&times;{10}^3& -3.598\&times;{10}^3\end{array}\right]}_{129\&times;101}$

By formula (13), obtaining Jacobi matrix C is:

$C={\left[\begin{array}{ccccc}0& ...& 1.7864& 1.4337& ...\\ 0& ...& -4.6070& 0.9400& ...\\ 0& ...& 0.9400& -4.4304& ...\\ 0& ...& 0.8712& 0.9486& ...\\ 0& ...& -3.2235& -2.7391& ...\\ 0& ...& 14.7412& -2.0712& ...\\ 0& ...& -2.0712& 14.5739& ...\\ 0& ...& -1.8608& -1.9540& ...\end{array}\right]}_{8\&times;101}$

(8) state variable upgrades

The correction amount x of computing mode variable ^(time), then upgrade state variable, obtain state variable and be newly worth; That is: x ^(time+1)=x ^(time)+ Δ x ^(time), time=time+1, correction computing formula is:

Δ x=(H ^tr ^-1h) ^-1{ H ^tr ^-1Δ z-C ^t[C (H ^tr ^-1h) ^-1c ^t] ^-1[C (H ^tr ^-1h) ^-1h ^tr ^-1Δ z-Δ c] } in (14) formula, time is for calculating iterations; H ^tand C ^tbe respectively the transposition of Jacobian matrix H and C; Δ c=-c (x); represent the measurement amount of unbalance that iterative value calculates based on formula (2)-(5) when being x, comprise injection that low-voltage distribution network platform region public affairs become three-phase node and neutral point in end points s each phase node Injection Current amplitude amount of unbalance and total active power amount of unbalance and total reactive power amount of unbalance, low-voltage network load subscriber endpoints i and to gain merit unbalanced power amount and inject amount of unbalance and the Voltage unbalance amount of reactive power; R ^-1for n rank unit matrix, the same step of its meaning (1).

With the citing of the result of the 1st iterative computation, by formula (14), calculate the correction amount x of state variable ⁽¹⁾for:

$\mathrm{\&Delta;}{x}^{\left(1\right)}={\left[\begin{array}{ccccccc}...& -0.1423& 0.0106& 0.6434& -0.1243& -0.390& ...\end{array}\right]}_{1\&times;101}^T$

(9) state estimation convergence judges

If the correction amount x of state variable ^(time)meet max (| Δ x ^(time)|) < ε, then finishing iteration calculates, and goes to step 3; If max (| Δ x ^(time)|)>=ε and iterations time≤Tmax, then return step 2, continue iteration and carry out state estimation; When max (| Δ x ^(time)|)>=ε and iterations time>Tmax, then directly stop iteration, export " state estimation does not restrain ".

With the citing of the result of the 1st iterative computation, now, ε=10 ^-4, Tmax=50, max (| Δ x ⁽¹⁾|)=229.1871> ε, time=1<Tmax.Judge according to convergence, return the step 2, again continue iteration and carry out state estimation.

According to above step 2, meet the condition of convergence after iteration 7 times, now max (| Δ x (7) |)=7.1394 × 10-5< ε, therefore done state estimates iterative computation, and go to step three.

Step 3: each phase active power and the reactive power of carrying out low-voltage distribution network platform region load subscriber endpoints are estimated;

Computing formula is:

In formula: with be respectively the active power of d phase power consumer and the estimated value of reactive power of low-voltage distribution network platform region load subscriber endpoints i; The same formula of remaining variables meaning (3).Its result of calculation is as shown in table 1 below:

Meritorious and the reactive power of table 1 power consumer measures estimated value

Step 4: each phase active power remote measurement evaluated error and the reactive power remote measurement evaluated error that calculate low-voltage distribution network platform region load subscriber endpoints;

Computing formula is:

$\left\{\begin{array}{c}{\mathrm{\&lambda;}}_{i,P}^d=|P_i^{\mathrm{dn}}-{\hat{P}}_i^{\mathrm{dn}}|/Z_i\\ {\mathrm{\&lambda;}}_{i,Q}^d=|Q_i^{\mathrm{dn}}-{\hat{Q}}_i^{\mathrm{dn}}|/Z_i\end{array}\right.,d\&Element;B_p---\left(16\right)$

In formula, with the d phase active power of low-voltage distribution network platform region load subscriber endpoints i and the remote measurement evaluated error of reactive power respectively; with same formula (3); with same formula (15); Z _ifor the active power of end points i or the measurement type reference value of reactive power, be for 380/220V low-voltage platform area electrical network herein, the measurement type reference value of active power and reactive power is all 145kW.Result of calculation is as shown in table 2 below:

Meritorious and the reactive power remote measurement evaluated error result of table 2 power consumer

Step 5: finally by comparing the active power of each phase power consumer of load end points and reactive power remote measurement evaluated error and state estimation qualification rate desired value size, determine whether suspicious stealing user.

The active power of described each phase power consumer by comparing load end points and reactive power remote measurement evaluated error and state estimation qualification rate desired value size, determine whether suspicious stealing user: if the d phase active power of low-voltage distribution network platform region load subscriber endpoints i or the remote measurement evaluated error of reactive power meet formula (17), then it is determined that the presence of stealing or exception electrical phenomena, the d phase power consumer namely picking out low-voltage distribution network platform region load subscriber endpoints i is suspicious stealing user; Otherwise normal electricity consumption user thought by the d phase power consumer of low-voltage distribution network platform region load subscriber endpoints i.

${\mathrm{\&lambda;}}_{i,P}^d\&GreaterEqual;2.5\%,\mathrm{or}{\mathrm{\&lambda;}}_{i,Q}^d\&GreaterEqual;2.5\%---\left(17\right)$

From the meritorious of phase power consumer each in table 2 and reactive power remote measurement evaluated error result, meritorious remote measurement evaluated error and the reactive power remote measurement evaluated error of all power consumers in platform district are all less than state estimation qualification rate, therefore, this district does not find suspicious stealing user, and all power consumers in Ji Gaitai district are judged to be it is normal electricity consumption user.

Experiment effect

With the IEEE13 node system shown in accompanying drawing 2 for object, the validity of citing checking the inventive method.For ease of the experiment effect of simple declaration the inventive method, as follows amendment is simplified to former IEEE-13 node system: with the earth and a phase node of public substation transformer end points for Zero potential reference, the neutral-point solid ground of Jin Tai district distribution low-voltage side point, branch road model is all set to 501, and spaced apart arranges identical, it is the spaced apart that this branch road a-c two is alternate.

To IEEE13 node regulation system, obtain network structure and parameter information, namely obtain platform district network line resistance, reactance and susceptance parameter, the rated voltage of circuit, power reference.And obtain that platform district distribution transforming end points 1 place measures that the three-phase of relative neutral point gathered always injects that meritorious, total injection is idle, three-phase Injection Current amplitude and voltage magnitude metric data, and the injection of the relative neutral point of all the other load end points place power consumers is meritorious, reactive power and voltage magnitude measure.Assuming that the b phase node load of the c phase load of end points 2 and end points 13 is stealing user, and the active power measuring value of b phase node load when abnormal electricity consumption of the c phase load and end points 13 that set end points 2 50% when being all reduced to normal electricity consumption, reactive power measuring value is reduced to during normal electricity consumption 75%.The fixed weight R of state estimation is set ^-1for unit battle array, state estimation convergence precision ε is 10 ^-4.Initialization maximum iteration time: Tmax=50, and make time=1.

First the basic data of IEEE13 update the system is inputted, and parameter initialization; Then according to the technical scheme in summary of the invention, the estimation of low-voltage distribution network three-phase state is carried out, max after iteration 8 times (| Δ x ⁽⁸⁾|)=2.5235 × 10 ^-6< ε, meet the state estimation condition of convergence, then go to step three, the estimated value of each meritorious and reactive power mutually calculating all load subscriber endpoints, and according to step 4, calculate the meritorious of all load users and reactive power remote measurement evaluated error, result of calculation is as shown in table 3 below.According to step 5, by the meritorious of load user and reactive power remote measurement evaluated error and state estimation qualification rate Indexes Comparison, judge whether the electric load user in this district's electrical network is suspicious stealing user one by one, as shown in Table 3, meritorious remote measurement evaluated error 20.173% and the idle remote measurement evaluated error 5.781% of the c phase power consumer of end points 2 are greater than state estimation qualification rate, gain merit remote measurement evaluated error 23.521% and reactive power remote measurement evaluated error 10.439% of the b phase node load of end points 13 is also all greater than state estimation qualification rate, and all the other node load power are all less than corresponding state estimation qualification rate, therefore, judge that the c phase electric load of end points 2 and the b phase node power load of end points 13 are as suspicious stealing user, this is the consistent of stealing user with the b phase power consumer of the c phase power consumer and end points 13 that set end points 2 before, namely identification and the location of stealing user is achieved, thus also demonstrate validity and the correctness of the inventive method.

Meritorious and the reactive power remote measurement evaluated error result of table 3 load electric power user

From experimental result:

1, the inventive method makes full use of the redundancy metric data that remote centralized meter-reading system gathers, and accurately can be calculated put into effect the meritorious of all load electric power users of district's electrical network and reactive power estimated value by state estimation;

2, the inventive method is by comparing telemetering of power evaluated error and state estimation qualification rate index size, and can go out suspicious stealing user by accurate recognition, identification effect is good;

3, the inventive method is by obtaining electric current and voltage and the power data of platform district electric network composition parameter and remote centralized meter-reading system collection, and utilize computer program just to realize the electrical network electricity consumption monitoring analysis of platform district exactly, method is simple, practical, easy to utilize.

The whole step process of the inventive method realizes automatically calculating result by computer programming.

The above, be only present pre-ferred embodiments, therefore can not limit scope of the invention process according to this, the equivalence change namely done according to the scope of the claims of the present invention and description with modify, all should still belong in scope that the present invention contains.

Claims (5)

1. the platform district electricity consumption monitoring method estimated based on low-voltage network three-phase state:

Step one: first input three-phase voltage, three-phase current, active power data and reactive power data and low-voltage distribution network platform region structured data that loaded termination gathers under section at one time, and init state estimated parameter information;

Step 2: carry out the estimation of low-voltage network three-phase state;

Step 3: each phase active power and the reactive power of carrying out low-voltage distribution network platform region load subscriber endpoints are estimated;

Step 4: each phase active power telemetering of power evaluated error and the reactive power remote measurement evaluated error that calculate low-voltage distribution network platform region load subscriber endpoints;

Step 5: finally by comparing the active power of each phase power consumer of load end points and reactive power remote measurement evaluated error and state estimation qualification rate desired value size, determine whether suspicious stealing user.

2. a kind of platform district electricity consumption monitoring method estimated based on low-voltage network three-phase state according to claim 1, it is characterized in that: described three-phase voltage, three-phase current and active power data and reactive power data comprise total active power under low-voltage distribution network platform region common transformer discontinuity surface in a period of time in office and total reactive power data and three-phase voltage amplitude and three-phase current amplitude, and each phase active power of low-voltage distribution network platform region all loads user terminal at one time under section and each phase reactive power data, three-phase voltage amplitude and phase current magnitude.

3. a kind of platform district electricity consumption monitoring method estimated based on low-voltage network three-phase state according to claim 2, is characterized in that: described low-voltage distribution network platform region structured data comprises: network line resistance, reactance and admittance parameter, the rated voltage of circuit, reference power.

4. a kind of platform district electricity consumption monitoring method estimated based on low-voltage network three-phase state according to claim 3, it is characterized in that: described init state estimated parameter information specifically refers to: the three-phase node voltage amplitude arranging all end points of low-voltage distribution network platform region is 1.0pu, and a phase voltage phase place is zero, b, c two-phase voltage-phase compared with a phase voltage delayed 120 ° and 240 ° respectively; Set condition is estimated to measure weight matrix R ^-1for n rank unit matrix, i.e. n rank unit matrix R ^-1diagonal entry be 1 entirely, off diagonal element to be 0, n be entirely measurement number actual in state estimation; Initialization maximum iteration time Tmax is 40 ~ 60; Convergence precision ε is 10 ^-3~ 10 ^-5, and iterations time=1 is set.

5. a kind of platform district electricity consumption monitoring method estimated based on low-voltage network three-phase state according to claim 4, is characterized in that: described low-voltage network three-phase state is estimated specifically to comprise the following steps:

1) computational grid bus admittance matrix;

2) each phase node Injection Current amplitude amount of unbalance of low-voltage distribution network platform region common transformer end points s and total active power amount of unbalance and total reactive power amount of unbalance is calculated;

3) each phase active power amount of unbalance and the reactive power amount of unbalance of the three-phase node of low-voltage distribution network platform region load subscriber endpoints i is calculated;

4) active power amount of unbalance and the reactive power amount of unbalance of the neutral point of low-voltage distribution network platform region load subscriber endpoints i is calculated;

5) amount of unbalance of the three-phase node voltage of low-voltage distribution network platform region load subscriber endpoints i is calculated;

6) the current imbalance amount of three-phase node and the current imbalance amount of neutral point in contact zero end points m is calculated;

7) measurement Jacobian matrix is calculated;

8) state variable upgrades: the correction amount x of computing mode variable ^(time), then upgrade state variable, obtain state variable and be newly worth;

9) state estimation convergence judges: if the correction amount x of state variable ^(time)meet max (| Δ x ^(time)|) < ε, then finishing iteration calculates, and goes to step three; If max (| Δ x ^(time)|)>=ε and iterations time≤Tmax, then return step 2, continue iteration and carry out state estimation; When max (| Δ x ^(time)|)>=ε and iterations time>Tmax, then directly stop iteration, export " state estimation does not restrain ".