CN101236216B - Electrical power system metric data time difference compensation state estimation method - Google Patents

Abstract

A power system measurement data time difference compensation state estimation method takes the quickest data acquisition time as the state estimation reference time to compensate measurement errors caused by other measurement time delays to obtain an updated electric network state. Regarding a PMU quick belt time mark measurement, updated real-time measurement data is adopted to take part in the state estimation and serve as the time of state estimation electric network section; regarding the measurements with time delays, corresponding time difference compensation factors are introduced according to the different time delays to finally obtain the updated state of the whole electric network through using the measurement variance of the updated continuous data section, modifying the time difference compensation factors according to the actual time delay conditions in the interactive process, and solving the time difference compensation state estimation equation.

Description

Electrical power system metric data time difference compensation state estimation method

One, technical field

The present invention relates to a kind of power system state estimation method, especially to the state estimation under the electric system measurement time delay condition difference.

Two, background technology

One megatrend of modern power systems development is that electrical network is interconnected, electrical network is interconnected when bringing great economic benefit, has also brought coordination control problem, has only from overall angle, electrical network is carried out synthesis optimizing and scheduling, could guarantee the security and the economy of interconnected network.Because adjacent electric network fault can directly have influence on the safety of internal electric network, the model and the data of adjacent electrical network are controlled no less important to electrical network analysis.Measure in order to obtain the whole network collection, often according to the different mining of communication condition with different transmission modes, different transmission modes have also caused the data of synchronization acquisition that the collection time difference is arranged.When setting up the interconnected network model, have only the time-delay of a few tens of milliseconds based on the measurement of GPS synchronous phasor measurement unit (PMU), but because its measurement has markers, can think there is not the time difference; The data that conventional remote terminal unit equipment (RTU) is gathered generally have several seconds time-delay; Remote terminal unit equipment (RTU), it is a kind of telemechanical apparatus that is installed in generating plant or transformer station in electrical network supervision and the control system, it is responsible for gathering the analog quantity and the quantity of state of reflection generating plant, place or converting station electric power running status, monitor and transmit these analog quantitys and quantity of state that the control and the traffic order of generating plant, place or transformer station mail at the operation dispatching center to the dispatching center.And will be a lot of slowly by the metric data of other control center's acquisition.Obtain data from other control center dual mode is generally arranged: a kind of is to transmit the remote measure and communication data in real time, and this mode generally has the 10-30 time-delay of second; Another kind of mode is after completion status is estimated in the EMS of other control center, the complete state estimation result of transmission, and this mode generally has the time-delay of a few minutes.After control center obtains adjacent electric network data, can in the state estimation of oneself, use these measurements or merge the state estimation result who obtains, thereby obtain a complete data section, use for safety assessment or Optimization Dispatching.In conventional state estimation, data error is considered to that measurement accuracy causes, also handles as error in measurement owing to gather the error that the time difference causes.And in interacted system, although be in the slow change procedure during electrical network stable state, the measurement of participating in calculating is a lot, because it is just very important to gather the error in measurement that the time difference causes, if do not take in, may produces the state estimation result and have a strong impact on.China's electrical network is carried out differentiated control, EMS at different levels system only safeguards electric network model and the image data in the own compass of competency, therefore, and outside setting up during pessimistic concurrency control, need obtain the real time data of outer net by the exchanges data between EMS, the time-delay of this exchanges data is very important too.

Under the electrical network normal running (operation) conditions, the time difference, upgrades comparatively fast if all measure owing to load variations causes the error that measure to produce, and consequent error is less, does not influence state estimation and provides correct electrical network section.When needs use the data of other control center's exchange, because it is bigger to measure time delay, there is bigger gap in borderline measurement between the different control zones, for the guaranteed output balance, conventional state estimation can be recognized as bad data to the load data of boundary vicinity, the Measurement Biases that the whole time difference is produced focuses on the border load, thereby cause the electric network state distortion, when situation is serious even can cause that state estimation does not restrain, even restrained, such result of calculation can not be as the foundation to dispatching of power netwoks control.

Can not fine processing in view of conventional state estimation owing to gather the error in measurement problem that the time difference brought, thereby press for a kind of new state estimation technology, this state estimation technology can take into full account the different collection time differences that measure, and obtains the electric network state more true, more consistent than conventional state estimation.

Three, summary of the invention

The present invention seeks to: for solve conventional method for estimating state can not fine processing owing to gather the error in measurement problem that the time difference brought, a kind of electrical power system metric data time difference compensation state estimation method has been proposed, measure data characteristic by measuring in conjunction with PMU measurement, conventional RTU, transmitting, give full play to the effect of rapid measuring, solved different measurement time-delays and caused the inconsistent problem of data section.Utilize method for estimating state of the present invention, can Fast estimation go out the fastest measurement collection electric network state constantly, and not need to wait for the measurement of time-delay maximum, and can adapt to the time delay condition of various variations.

Technical solution of the present invention is: electrical power system metric data time difference compensation state estimation method, and the fast speed belt markers that measures for PMU measures, and uses up-to-date real-time metric data to participate in state estimation, and as time of state estimation electrical network section; For the measurement that time-delay is arranged, difference according to time-delay, introduce the corresponding time difference compensation factor, utilize the measurement variable quantity and the time difference compensation factor of up-to-date continuous two data sections to revise the time-delay measurement by iteration, simultaneously in iterative process according to the actual time delay condition correction time difference compensation factor, by finding the solution time difference compensation state estimation equation, finally obtain the electric network state in the up-to-date moment of the whole network.

In conventional state estimation, can set up measurement equation to meritorious measurement and reactive voltage measurement respectively, as (1) formula.

$\left[\begin{array}{c}{Z}_p\\ Z_q\end{array}\right]=\left[\begin{array}{c}{h}_p(\mathrm{\θ},U)\\ h_q(\mathrm{\θ},U)\end{array}\right]+\left[\begin{array}{c}{v}_p(\mathrm{\θ},U)\\ v_q(\mathrm{\θ},U)\end{array}\right]---\left(1\right)$

Wherein, Z _pBe meritorious phasor, the Z of measuring _qFor reactive voltage measures phasor, h _p(θ U) is meritorious measurement equation phasor, h _q(θ U) is reactive voltage measurement equation phasor, v _p(θ U) is meritorious error in measurement phasor, v _q(θ, U) reactive voltage error in measurement phasor.θ is the busbar voltage angle, and U is the busbar voltage amplitude.

Set up meritorious and idle objective function J respectively _p(θ), J _q(θ), as (2), (3) formula.

$J_p\left(\mathrm{\θ}\right)={[Z_p-h_p(\mathrm{\θ},U)]}^T{R}_p^{-1}[Z_p-h_p(\mathrm{\θ},U)]---\left(2\right)$

$J_q\left(U\right)={[Z_q-h_q(\mathrm{\θ},U)]}^T{R}_q^{-1}[Z_q-h_q(\mathrm{\θ},U)]---\left(3\right)$

For given measurement phasor Z _p, Z _q, state phasor busbar voltage angle θ, amplitude U make objective function J exactly _p(θ), J _q(θ) be minimum θ, U value.After the linearization process, obtain PQ decoupling zero state estimation iteration correction formula (4) and (5) of basic weighted least-squares.

$\mathrm{\Δ\θ}={\left[H_p^T{R}_p^{-1}{H}_p\right]}^{-1}{H}_p^T{R}_p^{-1}[Z_p-h_p(\mathrm{\θ},U)]---\left(4\right)$

$\mathrm{\ΔU}={\left[H_q^T{R}_q^{-1}{H}_q\right]}^{-1}{H}_q^T{R}_q^{-1}[Z_q-h_q(\mathrm{\θ},U)]---\left(5\right)$

Measurement equation after the linearization is expressed as (6) formula.

$\left[\begin{array}{c}{Z}_p\\ Z_q\end{array}\right]=\left[\begin{array}{c}{H}_p\mathrm{\θ}\\ H_{q}U\end{array}\right]+\left[\begin{array}{c}{v}_p(\mathrm{\θ},U)\\ v_q(\mathrm{\θ},U)\end{array}\right]---\left(6\right)$

H wherein _pAnd H _qIt all is constant matrices.

If there are two sections, can set up measurement equation respectively as (7), (8) formula.

$\left[\begin{array}{c}{Z}_{p1}\\ Z_{q1}\end{array}\right]=\left[\begin{array}{c}{H}_p{\mathrm{\θ}}_1\\ H_q{U}_1\end{array}\right]+\left[\begin{array}{c}{v}_p({\mathrm{\θ}}_1,U_1)\\ v_q({\mathrm{\θ}}_1,U_1)\end{array}\right]---\left(7\right)$

$\left[\begin{array}{c}{Z}_{P2}\\ Z_{q2}\end{array}\right]=\left[\begin{array}{c}{H}_P{\mathrm{\θ}}_2\\ H_q{U}_2\end{array}\right]+\left[\begin{array}{c}{v}_P({\mathrm{\θ}}_2,U_2)\\ v_q({\mathrm{\θ}}_2,U_2)\end{array}\right]---\left(8\right)$

Two section measurement equations subtract each other, and obtain the section variation measurement equation as (9) formula.

$\left[\begin{array}{c}{Z}_{\mathrm{\Δp}}\\ Z_{\mathrm{\Δq}}\end{array}\right]=\left[\begin{array}{c}{Z}_{p2}-Z_{P1}\\ Z_{q2}-Z_{q1}\end{array}\right]=\left[\begin{array}{c}{H}_p{\mathrm{\θ}}_2\\ H_q{U}_2\end{array}\right]-\left[\begin{array}{c}{H}_p{\mathrm{\θ}}_1\\ H_q{U}_1\end{array}\right]+\left[\begin{array}{c}{v}_p({\mathrm{\θ}}_2,U_2)\\ v_q({\mathrm{\θ}}_2,U_2)\end{array}\right]-\left[\begin{array}{c}{v}_p({\mathrm{\θ}}_1,U_1)\\ v_q({\mathrm{\θ}}_1,U_1)\end{array}\right]---\left(9\right)$

$=\left[\begin{array}{c}{H}_p{\mathrm{\θ}}_{\mathrm{\Δ}}\\ H_q{U}_{\mathrm{\Δ}}\end{array}\right]+\left[\begin{array}{c}{v}_p({\mathrm{\θ}}_2,U_2)\\ v_q({\mathrm{\θ}}_2,U_2)\end{array}\right]-\left[\begin{array}{c}{v}_p({\mathrm{\θ}}_1,U_1)\\ v_q({\mathrm{\θ}}_1,U_1)\end{array}\right]$

Can obtain weighted least-squares state variation amount iterative formula:

$\mathrm{\Δ}{\mathrm{\θ}}_{\mathrm{\Δ}}={\left[H_p^T{R}_p^{-1}{H}_p\right]}^{-1}{H}_p^T{R}_p^{-1}[Z_{\mathrm{\Δp}}-H_p{\mathrm{\θ}}_{\mathrm{\Δ}}]---\left(11\right)$

$\mathrm{\Δ}{U}_{\mathrm{\Δ}}={\left[H_q^T{R}_q^{-1}{H}_q\right]}^{-1}{H}_q^T{R}_q^{-1}[Z_{\mathrm{\Δq}}-H_q{U}_{\mathrm{\Δ}}]---\left(12\right)$

As seen, be the same based on the state estimation information matrix of state variation amount with conventional state estimation, but error in measurement is calculated more simple in its iterative process.

In conventional state estimation, introduce the time difference compensation factor, the compensating factor initial value is zero, and participates in the state estimation iterative computation as quantity of state, according to the corresponding measurement actual time delay condition correction-compensation factor.

By the variable quantity state estimation, obtained the variation of each measuring point, thereby can utilize variable quantity that the measuring point that the time difference is arranged is compensated.Revise the measurement equation of state estimation, introduce the time difference compensation factor as (13) formula

$\left[\begin{array}{c}{Z}_p+{\mathrm{\φ}}_p{P}_{\mathrm{\Δ}}\\ Z_q+{\mathrm{\φ}}_q{Q}_{\mathrm{\Δ}}\end{array}\right]=\left[\begin{array}{c}{h}_p(\mathrm{\θ},U)\\ h_q(\mathrm{\θ},U)\end{array}\right]+\left[\begin{array}{c}{v}_p(\mathrm{\θ},U)\\ v_q(\mathrm{\θ},U)\end{array}\right]---\left(13\right)$

Wherein, P _ΔBe meritorious measuring point state variation estimated value, Q _ΔBe reactive voltage measuring point state variation estimated value, φ is the time difference compensation factor,

Δ t ₀Be the interval time of two continuous cross-sections, Δ t ₁The time difference that need compensate with respect to the benchmark measurement time for certain measuring point.After introducing compensating factor, just the time interval to the continuous acquisition metric data does not have uniform requirement, has time-delay to measure and can use up-to-date image data, does not need accurately to know simultaneously to measure the actual time difference, has simplified the measurement consistance and has found the solution complexity.Consider that the meritorious idle measurement of same equipment has the same time difference, can take the same time difference compensation factor, suc as formula (14).

$\left[\begin{array}{c}{Z}_p\\ Z_q\end{array}\right]=\left[\begin{array}{c}{h}_p(\mathrm{\θ},U)-\mathrm{\φ}{Z}_{\mathrm{\Δp}}\\ h_q(\mathrm{\θ},U)-\mathrm{\φ}{Z}_{\mathrm{\Δq}}\end{array}\right]+\left[\begin{array}{c}{v}_p(\mathrm{\θ},U)\\ v_q(\mathrm{\θ},U)\end{array}\right]---\left(14\right)$

Formula (14) is carried out linearization process, can obtain weighted least-squares time difference compensation state estimation formula:

$\left[\begin{array}{c}\mathrm{\Δ\θ}\\ \mathrm{\Δ\φ}\end{array}\right]={\left[N_p^T{R}_p^{-1}{N}_p\right]}^{-1}{N}_p^T{R}_p^{-1}[Z_p-h_p(\mathrm{\θ},U)+\mathrm{\φ}{P}_{\mathrm{\Δ}}]---\left(15\right)$

$\left[\begin{array}{c}\mathrm{\ΔU}\\ \mathrm{\Δ\φ}\end{array}\right]={\left[N_q^T{R}_q^{-1}{N}_q\right]}^{-1}{N}_q^T{R}_q^{-1}[Z_q-h_q(\mathrm{\θ},U)+\mathrm{\φ}{Q}_{\mathrm{\Δ}}]---\left(16\right)$

N wherein _p, N _qBe respectively meritorious idle measurement augmented matrix, wherein

E _p, E _qBe the matrix of a mxc dimension, m is meritorious or idle measurement number, the time difference compensation factor number of c for introducing, and whenever walk to and have only an element more, and value is for-1.Adopt time difference compensation, need be with a measuring point acquisition time as reference time, the time difference compensation factor that measure reference time is zero, corresponding E matrix one provisional capital is a neutral element.E matrix one provisional capital is zero pseudo-measurement of zero injection of node in addition.According to measuring the collection characteristics, one group of measurement can be used the same time difference compensation factor, is unit with the factory station for example, to one group of metric data under the identical acquisition mode, set its identical time difference compensation factor, can significantly reduce the number of finding the solution the time difference compensation factor like this.For mode according to complete section swap data, can set a time difference compensation factor to whole profile data, increase the just effectively influence of treatment time difference generation of calculated amount seldom like this.

In order to improve the practicality of time difference compensation state estimation, can not use the variable quantity state estimation, and two section variation amounts that directly adopt each measuring point to measure, the time difference compensation state estimation only need once be calculated like this, and computing velocity is suitable with conventional method for estimating state.

To the measurement of time-delay is arranged, introduce two continuous data section shape attitudes and change, metric data is expressed as Z+ φ Z in measurement equation _Δ, wherein Z is up-to-date measurement, φ measures compensating factor, Z for time-delay is arranged _ΔBe continuous two the section variation amounts of electrical network.For improving computing velocity and adaptability, continuous two data section variation can adopt the variable quantity of continuous two measurements, avoid carrying out the variable quantity state estimation.

The invention has the beneficial effects as follows: in conjunction with the characteristics that PMU measures, conventional RTU measures, the effect of performance rapid measuring has solved different measurement time-delays and has caused the inconsistent problem of data section.Can Fast estimation go out the fastest measurement collection electric network state constantly, and not need to wait for the measurement of time-delay maximum, and can adapt to the time delay condition of various variations.It is little to measure time difference corrected Calculation amount in the iterative process, and speed is fast, and computation process is simple.Can be simultaneously in iterative process according to the actual time delay condition correction time difference compensation factor.

Four, description of drawings

Fig. 1 is the simple wiring schematic diagram of one-of-a-kind system, and the remote measurement at circuit top factory station is Pa, and terminal remote measurement is Pb.

Fig. 2 is that meritorious measurement of circuit changes synoptic diagram.

Five, embodiment

Shown in Figure 1, if ignore the line power loss, then: Pb ≈-Pa, if Pa remote measurement value according to linear change, Pb is also according to linear change so.Constantly gathered the data section respectively at t2, t4, if the Pb measuring point has time-delay, Pb measuring point actual acquisition is t1, t3 power constantly so, as shown in Figure 2.

Fig. 2 line power measures shown in the variation, does not have the time difference if gather, and the measurement during the state variation amount is estimated should be Pa4-Pa2 and Pb4-Pb2, can obtain variance estimation.Suppose that there is constant time lag in the Pb collection, because t4-t2=t3-t1, then Pb4-Pb2=Pb3-Pb1.As seen from the above analysis, when adopting the state variation amount to estimate, can directly adopt the measurement difference of two sections to form the measurement phasor, the collection time difference of each measuring point is little to the result of calculation influence.

If adopt conventional method for estimating state as can be seen from the example of Fig. 2, when t4 carries out state estimation constantly, measure and be respectively Pa4 and Pb3, the measurement time difference of Pb measuring point has been caused the state estimation error like this, even the estimated result that can lead to errors.Therefore the measurement of Pb3 need be carried out time difference compensation, and to obtain t4 Pb4 constantly, result of calculation just may be correct like this.

This shows, if electrical network in slow change procedure, each point measures can think linear change within a short period of time, that is to say, as long as it is identical to measure the acquisition interval time, its measurement changes also approximately equal.

CHINA SOUTHERN POWER is by the model merging formation above electric network model of the 220kV that comprises Guizhou, Yunnan, Guangxi, Guangdong, and data acquisition modes comprises that straight accent station (mainly being 500kV transformer station and large power plant) RTU of factory directly adopts, and each is economized and transfers EMS forwarding state estimated result.Because economizing, each transfer a few minutes to transmit a state estimation result, province turn send out data and directly adopt the data time difference bigger, the data of economizing between transferring also have than large time difference, the inconsistent problem of data is more serious, influenced the state estimation convergence, unavailable in state estimation result of calculation of very fast period of load variations.

Adopt the time difference compensation computing method, be respectively Guizhou, Yunnan, Guangxi, the Guangdong forwarding data introducing time difference compensation factor, increased by 4 quantity of states altogether, that is: think that RTU directly adopts data and do not have the time difference, and as the reference data section.Turn the up-to-date profile data sent out and the difference of a last profile data is participated in state estimation as measuring variable quantity with each province, time difference compensation factor initial value is made as 0, and revises in iterative process.

After adopting the time difference compensation algorithm, time difference compensation factor calculated value is generally in 0 to 1 scope, state estimation computing time and conventional method for estimating state difference are very little, all in 1.5 seconds, the state estimation convergency factor has been brought up to about 3%, calculate as image data according to the measurement behind the time difference compensation, state estimation measures qualification rate and has generally improved about 2%, reduced the error of directly adopting data and transmitting data boundary simultaneously, more near the actual electric network state, improved the availability of state estimation result of calculation in the result of calculation of load variations period.

Claims (5)

1. electrical power system metric data time difference compensation state estimation method, it is characterized in that with the fastest image data time as state estimation reference time, compensate other and measure the error in measurement that time-delay produces, to obtain up-to-date electric network state; Measure for the fast speed belt markers of PMU, use up-to-date real-time metric data to participate in state estimation, and as time of state estimation electrical network section; For the measurement that time-delay is arranged, difference according to time-delay, introduce the corresponding time difference compensation factor, utilize the measurement variable quantity of up-to-date continuous data section, simultaneously in iterative process according to the actual time delay condition correction time difference compensation factor, by finding the solution time difference compensation state estimation equation, finally obtain the electric network state in the up-to-date moment of the whole network;

In conventional state estimation, respectively meritorious measurement and reactive voltage are measured measurement equation (1) formula of setting up:

$\left[\begin{array}{c}{Z}_p\\ Z_q\end{array}\right]=\left[\begin{array}{c}{h}_p(\mathrm{\θ},U)\\ h_q(\mathrm{\θ},U)\end{array}\right]+\left[\begin{array}{c}{v}_p(\mathrm{\θ},U)\\ v_q(\mathrm{\θ},U)\end{array}\right]---\left(1\right)$

Wherein, Z _pBe meritorious phasor, the Z of measuring _qFor reactive voltage measures phasor, h _p(θ U) is meritorious measurement equation phasor, h _q(θ U) is reactive voltage measurement equation phasor, v _p(θ U) is meritorious error in measurement phasor, v _q(θ, U) reactive voltage error in measurement phasor; θ is the busbar voltage angle, and U is the busbar voltage amplitude;

Set up meritorious and idle objective function J respectively _p(θ), J _q(θ), (2), (3) formula:

$J_p\left(\mathrm{\θ}\right)={[Z_p-h_p(\mathrm{\θ},U)]}^T{R}_p^{-1}[Z_p-h_p(\mathrm{\θ},U)]---\left(2\right)$

$J_q\left(U\right)={[Z_q-h_q(\mathrm{\θ},U)]}^T{R}_q^{-1}[Z_q-h_q(\mathrm{\θ},U)]---\left(3\right)$

For given measurement phasor Z _p, Z _q, state phasor busbar voltage angle θ, amplitude U make objective function J exactly _p(θ), J _q(θ) be minimum θ, U value; After the linearization process, obtain PQ decoupling zero state estimation iteration correction formula (4) and (5) of basic weighted least-squares;

$\mathrm{\Δ\θ}={\left[H_p^T{R}_p^{-1}{H}_p\right]}^{-1}{H}_p^T{R}_p^{-1}[Z_p-h_p(\mathrm{\θ},U)]---\left(4\right)$

$\mathrm{\ΔU}={\left[H_q^T{R}_q^{-1}{H}_q\right]}^{-1}{H}_q^T{R}_q^{-1}[Z_q-h_q(\mathrm{\θ},U)]---\left(5\right)$

Measurement equation after the linearization is expressed as (6) formula, wherein H _pAnd H _qIt all is constant matrices;

$\left[\begin{array}{c}{Z}_p\\ Z_q\end{array}\right]=\left[\begin{array}{c}{H}_p\mathrm{\θ}\\ H_{q}U\end{array}\right]+\left[\begin{array}{c}{v}_p(\mathrm{\θ},U)\\ v_q(\mathrm{\θ},U)\end{array}\right]---\left(6\right)$

If there are two sections, set up measurement equation (7), (8) formula respectively:

$\left[\begin{array}{c}{Z}_{p1}\\ Z_{q1}\end{array}\right]=\left[\begin{array}{c}{H}_p{\mathrm{\θ}}_1\\ H_q{U}_1\end{array}\right]+\left[\begin{array}{c}{v}_p({\mathrm{\θ}}_1,U_1)\\ v_q({\mathrm{\θ}}_1,U_1)\end{array}\right]---\left(7\right)$

$\left[\begin{array}{c}{Z}_{P2}\\ Z_{q2}\end{array}\right]=\left[\begin{array}{c}{H}_P{\mathrm{\θ}}_2\\ H_q{U}_2\end{array}\right]+\left[\begin{array}{c}{v}_P({\mathrm{\θ}}_2,U_2)\\ v_q({\mathrm{\θ}}_2,U_2)\end{array}\right]---\left(8\right)$

Two section measurement equations subtract each other, and obtain section variation measurement equation (9) formula:

$\left[\begin{array}{c}{Z}_{\mathrm{\Δp}}\\ Z_{\mathrm{\Δq}}\end{array}\right]=\left[\begin{array}{c}{Z}_{p2}-Z_{P1}\\ Z_{q2}-Z_{q1}\end{array}\right]=\left[\begin{array}{c}{H}_p{\mathrm{\θ}}_2\\ H_q{U}_2\end{array}\right]-\left[\begin{array}{c}{H}_p{\mathrm{\θ}}_1\\ H_q{U}_1\end{array}\right]+\left[\begin{array}{c}{v}_p({\mathrm{\θ}}_2,U_2)\\ v_q({\mathrm{\θ}}_2,U_2)\end{array}\right]-\left[\begin{array}{c}{v}_p({\mathrm{\θ}}_1,U_1)\\ v_q({\mathrm{\θ}}_1,U_1)\end{array}\right]$

(9)

$=\left[\begin{array}{c}{H}_p{\mathrm{\θ}}_{\mathrm{\Δ}}\\ H_q{U}_{\mathrm{\Δ}}\end{array}\right]+\left[\begin{array}{c}{v}_p({\mathrm{\θ}}_2,U_2)\\ v_q({\mathrm{\θ}}_2,U_2)\end{array}\right]-\left[\begin{array}{c}{v}_p({\mathrm{\θ}}_1,U_1)\\ v_q({\mathrm{\θ}}_1,U_1)\end{array}\right]$

Obtain weighted least-squares state variation amount iterative formula:

${\mathrm{\Δ\θ}}_{\mathrm{\Δ}}={\left[H_p^T{R}_p^{-1}{H}_p\right]}^{-1}{H}_p^T{R}_p^{-1}[Z_{\mathrm{\Δp}}-H_p{\mathrm{\θ}}_{\mathrm{\Δ}}]---\left(11\right)$

${\mathrm{\ΔU}}_{\mathrm{\Δ}}={\left[H_q^T{R}_q^{-1}{H}_q\right]}^{-1}{H}_q^T{R}_q^{-1}[Z_{\mathrm{\Δq}}-H_q{U}_{\mathrm{\Δ}}]---\left(12\right)$

The state estimation information matrix of base state variation amount is the same with conventional state estimation;

In the conventional state estimation, introduce the time difference compensation factor, the compensating factor initial value is zero, and participates in the state estimation iterative computation as quantity of state, according to the corresponding measurement actual time delay condition correction-compensation factor; Revise the measurement equation of state estimation, introduce the time difference compensation factor as (13) formula:

$\left[\begin{array}{c}{Z}_p+{\mathrm{\φ}}_p{P}_{\mathrm{\Δ}}\\ Z_q+{\mathrm{\φ}}_q{Q}_{\mathrm{\Δ}}\end{array}\right]=\left[\begin{array}{c}{h}_p(\mathrm{\θ},U)\\ h_q(\mathrm{\θ},U)\end{array}\right]+\left[\begin{array}{c}{v}_p(\mathrm{\θ},U)\\ v_q(\mathrm{\θ},U)\end{array}\right]---\left(13\right)$

Wherein, P _ΔBe meritorious measuring point state variation estimated value, Q _ΔBe reactive voltage measuring point state variation estimated value, φ is the time difference compensation factor,

Δ t ₀Be the interval time of two continuous cross-sections, Δ t ₁The time difference that need compensate with respect to the benchmark measurement time for certain measuring point.

2. according to the described electrical power system metric data time difference compensation state estimation method of claim 1, after it is characterized in that introducing compensating factor, use up-to-date image data, formula (14) for the measurement that time-delay is arranged:

$\left[\begin{array}{c}{Z}_p\\ Z_q\end{array}\right]=\left[\begin{array}{c}{h}_p(\mathrm{\θ},U)-{\mathrm{\φZ}}_{\mathrm{\Δp}}\\ h_q(\mathrm{\θ},U)-{\mathrm{\φZ}}_{\mathrm{\Δq}}\end{array}\right]+\left[\begin{array}{c}{v}_p(\mathrm{\θ},U)\\ v_q(\mathrm{\θ},U)\end{array}\right]\text{---}\left(14\right)$

Formula (14) is carried out linearization process, obtains weighted least-squares time difference compensation state estimation iterative formula:

$\left[\begin{array}{c}\mathrm{\Δ\θ}\\ \mathrm{\Δ\φ}\end{array}\right]={\left[N_p^T{R}_p^{-1}{N}_p\right]}^{-1}{N}_p^T{R}_p^{-1}[Z_p-h_p(\mathrm{\θ},U)+{\mathrm{\φP}}_{\mathrm{\Δ}}]---\left(15\right)$

$\left[\begin{array}{c}\mathrm{\ΔU}\\ \mathrm{\Δ\φ}\end{array}\right]={\left[N_q^T{R}_q^{-1}{N}_q\right]}^{-1}{N}_q^T{R}_q^{-1}[Z_q-h_q(\mathrm{\θ},U)+{\mathrm{\φQ}}_{\mathrm{\Δ}}]---\left(16\right)$

N wherein _p, N _qBe respectively meritorious idle measurement augmented matrix, wherein

E _p, E _qBe the matrix of a m * c dimension, m is meritorious or idle measurement number, the time difference compensation factor number of c for introducing, and whenever walk to and have only an element more, and value is for-1;

Adopt time difference compensation, as reference time, the time difference compensation factor that measure reference time is zero with a measuring point acquisition time, and corresponding E matrix one provisional capital is a neutral element; E matrix one provisional capital is zero pseudo-measurement of zero injection of node in addition; According to measuring the collection characteristics, one group measures the same time difference compensation factor of use, to one group of metric data under the identical acquisition mode, sets its identical time difference compensation factor, finds the solution the number of the time difference compensation factor in order to minimizing.

3. electrical power system metric data time difference compensation state estimation method according to claim 1 is characterized in that for the mode according to complete section swap data, and whole profile data is only set a time difference compensation factor, increases calculated amount seldom like this.

4. electrical power system metric data time difference compensation state estimation method according to claim 1, it is characterized in that in order to improve the practicality of time difference compensation state estimation, and two section variation amounts that directly adopt each measuring point to measure, the time difference compensation state estimation only need once be calculated like this, and computing velocity is suitable with conventional method for estimating state.

5. electrical power system metric data time difference compensation state estimation method according to claim 1 is characterized in that the measurement of time-delay is arranged, and introduces two continuous data section shape attitudes and changes, and metric data is expressed as Z+ φ Z in measurement equation _Δ, wherein Z is up-to-date measurement, φ is the time difference compensation factor, Z _ΔBe continuous two the section variation amounts of electrical network; For improving computing velocity and adaptability, continuous two data section variation adopt the variable quantity that measures continuously, avoid the variable quantity state estimation.

Images