CN103001218A - Reactive voltage coordination control method based on comprehensive benefits of transmission and distribution power grid - Google Patents

Abstract

The invention provides a reactive voltage coordination control method based on comprehensive benefits of a transmission and distribution power grid. The reactive voltage coordination control method comprises enabling a subordinate power grid to upload gate characteristics to a superior control center; enabling the superior control center to establish a superior and subordinate reactive voltage coordinated optimization global model according to the gate characteristics; performing virtual control partitioning on a superior administered power grid, and obtaining a coordinated optimization calculation model of a plurality of virtual control partitions; enabling the superior control center to obtain optimized values of gate coordination variables of the subordinate power grid by calculation according to the coordinated optimization calculation model and send the optimized values to the subordinate power grid; enabling the subordinate power grid to add coordination variable restraint in an automatic voltage control (AVC) algorithm according to the optimized values of the gate coordination variables, and obtaining a virtual control partition optimization strategy based on the comprehensive benefits of the transmission and distribution power grid; and enabling the superior power grid and the subordinate power grid to perform reactive voltage coordination control according to the virtual control partition optimization strategy. The method improves the running state of the whole network, improves system benefits, reduces model calculation scales and complexity, and is good in reliability and high in compatibility.

Description

Reactive voltage control method for coordinating based on the transmission and distribution network comprehensive benefit

Technical field

The present invention relates to power system voltage control technology field, particularly a kind of reactive voltage control method for coordinating based on the transmission and distribution network comprehensive benefit.

Background technology

In power system reactive power voltage control field, different scheduling systems has determined the executor of interregional coordination control in the transmission and distribution network.In layering and zoning scheduling system, coordinate control and carried out by higher level control centre, critical point set point or interval that subordinate's electrical network issues by receiving higher level control centre, decoupling zero control between feasible region.Higher level control centre is not only as Consultation Center, and electric power networks and the corresponding controllable resources of own administration also can be arranged simultaneously, for example, in China's electrical network system, economizes that to transfer be exactly the higher level that transfer in each time possession, the equipment that the while also administers to some extent self.In order to guarantee network-wide security, high-quality, economical operation, higher level control centre also needs to take into full account regulating power, demand for control and the benefit etc. of subordinate's electrical network when the regional power grid of administering is optimized and controls.

For example, economize ground AVC (Automatic Voltage Control, automatism voltage control) coordinates to belong to the skimble-scamble situation of control centre of the superior and the subordinate information modeling, introduce subordinate's electrical network full detail if economize regulation and control center processed, then the real time data amount is too huge, therefore general is equivalent to load bus to subordinate's electrical network, subordinate's electrical network is not carried out complete modeling.If only have the imperfect information of subordinate's electrical network will be difficult to from the whole network angle calculation optimal load flow, thereby can't obtain each critical point specification of variables value.Therefore traditional way only makes province that ground two-stage operation of power networks is in reasonable level.

In order to guarantee network-wide security, high-quality, economical operation, power transmission network side control centre is when being optimized and controlling institute's compass of competency electrical network, the ability, demand and the benefit that need consideration and the skimble-scamble subordinate of its information modeling electrical network, and coordinate, and prior art can not well address the above problem.

Summary of the invention

The present invention is intended to solve at least one of technical problem that exists in the prior art.For this reason, the object of the invention is to propose a kind of reactive voltage control method for coordinating based on the transmission and distribution network comprehensive benefit.This method has been improved the whole network running status, has improved system benefit, has reduced model calculation scale and complexity, and good reliability, and is compatible high.

For achieving the above object, embodiments of the invention have proposed a kind of reactive voltage control method for coordinating based on the transmission and distribution network comprehensive benefit, may further comprise the steps: electrical network superior control centre of subordinate uploads the critical point characteristic, wherein, described critical point characteristic comprises regulating power, demand for control and optimization benefit; Described higher level control centre is converted into coordination variable constraint and the target correction term that characterizes with the critical point variable with described critical point characteristic, sets up reactive voltage coordination optimization world model of the superior and the subordinate; According to reactive voltage coordination optimization world model of described the superior and the subordinate, described higher level is administered electrical network carry out the virtual controlling subregion, obtain the coordination optimization computation model of a plurality of virtual controlling subregions; Described higher level control centre is according to described coordination optimization computation model, calculate described subordinate electrical network the critical point coordination variable optimal value and be sent to described subordinate electrical network; Described higher level's electrical network and described subordinate electrical network carry out reactive voltage according to described virtual controlling partition zone optimizing strategy and coordinate control.

Consider the critical point characteristic of subordinate's electrical network according to the embodiment of the invention based on the reactive voltage control method for coordinating of transmission and distribution network comprehensive benefit, higher level's power transmission network is carried out control partition, and computation model and the Practical resolving Algorithm of proposition control partition, improve the running status of the whole network, improved the system benefit of the whole network.By the superior and the subordinate's electric network reactive-load voltage coordination problem in the control partition that is dimensionality reduction with large electric network reactive-load voltage Harmonic Control decomposition and inversion, reduce computation model scale and computation complexity, and improved the compatibility of the benefit expression-form that traditional algorithm uploads subordinate, reduced the extra calculation cost of subordinate electrical network.

In one embodiment of the invention, but described regulating power be used for representing that described subordinate electrical network is to the idle modification scope in described critical point.

In one embodiment of the invention, it is qualified that described demand for control is used for controlling described subordinate line voltage, and shown in the critical point voltage restriction range that needs the higher level to control.

In one embodiment of the invention, described optimization benefit is when the described virtual controlling partition zone optimizing strategy of control does not clash on benefit, each subordinate's grid benefit of considering in the coordination optimization computation model.

In one embodiment of the invention, described optimization benefit is according to the electrical network AVC of subordinate, take described critical point variable as independent variable, by curve form or the statement of interval form.Use as the case may be different method for expressing, it is multi-form to have considered that subordinate's network optimization benefit is uploaded, and has improved subordinate is uploaded the compatibility of optimizing the benefit expression-form

In one embodiment of the invention, also comprise the improvement algorithm, be used for the virtual controlling subregion is optimized calculating, integrate and generate new virtual partition that wherein, the number of described new virtual partition is less than the virtual partition number before optimizing.With the superior and the subordinate's electric network reactive-load voltage coordination problem in the large electric network reactive-load voltage Harmonic Control decomposition and inversion control partition that is dimensionality reduction, and further virtual partition is optimized, reduced computation model scale and computation complexity, simultaneously subregion decoupling zero control has improved the reliability of AVC algorithm.If certain subregion occurs unusual, generally can not have influence on the normal coordination optimization control of other subregions yet.

In one embodiment of the invention, described higher level control centre is optimized calculating, obtains each control partition VACE set point.

In one embodiment of the invention, described the virtual controlling subregion is optimized calculating, integrates and generate new virtual partition and further comprise:

Step 1: select constraint subspace S _In' begin to calculate;

Step 2: making the described constraint subspace of current calculating is S _Ij', find the solution following computation model:

min f _H(x',u',Q _Li,U _Hi)

s.t.g _i(x',u',U _Hi,Q _Li)=0

h _j(x',u',U _Hi,Q _Li)≤0

x' _min<x'<x' _max

u' _min<u'<u' _max

Q _Lij∈S' _ij=[l' _ij,u' _ij]i=1,2,…,N,1≤j≤n

U _Hi，min<U _Hi<U _Hi，max i＝1,2,…,N'

-δ≤I _i,VACE+C _i,VACEΔu' _i≤δ

Wherein, x ', u ' are the subset of higher level's electric network state amount x and controlled quentity controlled variable u, and U is described critical point voltage, and Q is that described critical point is idle, and min and max represent respectively corresponding minimum and maximum threshold value, I _{I, VACE}Be the VACE index of described constraint subspace i, C _{I, VACE}Be the control sensitivity of described constraint subspace i to VACE, Δ u _i' be control deviation, δ is the VACE maximum deviation that regional voltage control system allows.Calculate locally optimal solution f _{Ij, opt}', can obtain coordinating subarea i constraint subspace S _In' locally optimal solution drop point x _Ij' and locally optimal solution f _{HL, ij, opt}, wherein,

f _HL,ij,opt=f' _ij,opt+f' _ij

Step 3: if drop point x _IjBelong to other subspaces S _Ik', basis then

Relation, obtain subspace S _Ik' locally optimal solution f _{HL, ik, opt}, wherein, f _{HL, ik, opt}=f' _{Ij, opt}+ f ' _Ik

If described drop point x _Ij' belong to a plurality of subspaces, then the interval numbering that interval rank numbering is minimum is designated as N _TEMPBecause

F ' _Ik＜f' _Ij, described interval rank numbering N _TEMPLocally optimal solution f _{HL, ik, opt}Be the relatively minimum solution of the network loss in described a plurality of spaces.

Step 4: if calculate for the first time, remember described interval rank numbering N _TEMPDescribed locally optimal solution f _{HL, ik, opt}Be f _{HL, i, temp}, and remember current drop point x _Ij' be x _{I, temp}Once calculated more described f if carried out _{HL, ik, opt}With described f _{HL, i, temp}If satisfy f _{HL, ik, opt}＜f _{HL, i, temp}, then make f _{HL, i, temp}=f _{HL, ik, opt}, x _{I, temp}=x _Ij'; Otherwise, if satisfy f _{HL, i, temp}＜f' _{Ij, opt}+ f _I1', then improve algorithm iteration and finish.

Step 5: if described interval numbering N _TEMP+ 1 is non-vanishing, gets corresponding described constraint subspace

Execution in step 2.

If described interval numbering N _TEMP+ 1 is zero, then improves algorithm iteration and finishes.

Step 6: single described control partition i calculates and finishes, and obtains the described optimal solution f of described control partition i _{HL, i, opt}With optimal solution drop point x _i, wherein, f _{HL, i, opt}=f _{HL, i, temp}, x _i=x _{I, temp}

According to the embodiment of the invention.。。。。

Additional aspect of the present invention and advantage in the following description part provide, and part will become obviously from the following description, or recognize by practice of the present invention.

Description of drawings

Above-mentioned and/or additional aspect of the present invention and advantage are from obviously and easily understanding becoming the description of embodiment in conjunction with following accompanying drawing, wherein:

Fig. 1 is the flow chart based on the reactive voltage control method for coordinating of transmission and distribution network comprehensive benefit according to the embodiment of the invention;

Fig. 2 is that the corresponding control partition of IEEE39 node system that adopts in the simulation example is divided schematic diagram

Fig. 3 is total active power loss contrast schematic diagram of the system of simulation example total system algorithms of different when being node load or burden without work sustainable growth situation;

The power transmission network active power loss of algorithms of different contrast schematic diagram when Fig. 4 is node load or burden without work sustainable growth situation;

The total active power loss contrast of the power distribution network of algorithms of different schematic diagram when Fig. 5 is node load or burden without work sustainable growth situation;

The subsystem A power transmission network side active power loss of algorithms of different contrast schematic diagram when Fig. 6 is node load or burden without work sustainable growth situation;

The subsystem B power transmission network side active power loss of algorithms of different contrast schematic diagram during Fig. 7 node load or burden without work sustainable growth situation;

No. 1 power distribution network active power loss contrast schematic diagram of algorithms of different when Fig. 8 is node load or burden without work sustainable growth situation;

No. 2 power distribution network active power loss contrast schematic diagrames of algorithms of different when Fig. 9 is node load or burden without work sustainable growth situation;

No. 1 power distribution network critical point voltage-contrast schematic diagram of algorithms of different when Figure 10 is node load or burden without work sustainable growth situation;

No. 2 power distribution network critical point voltage-contrast schematic diagrames of algorithms of different when Figure 11 is node load or burden without work sustainable growth situation;

The idle contrast schematic diagram in No. 1 power distribution network critical point of algorithms of different when Figure 12 is node load or burden without work sustainable growth situation; With

The idle contrast schematic diagram in No. 2 power distribution network critical points of algorithms of different when Figure 13 is node load or burden without work sustainable growth situation.

Embodiment

The below describes embodiments of the invention in detail, and the example of described embodiment is shown in the drawings, and wherein identical or similar label represents identical or similar element or the element with identical or similar functions from start to finish.Be exemplary below by the embodiment that is described with reference to the drawings, only be used for explaining the present invention, and can not be interpreted as limitation of the present invention.

Below with reference to the reactive voltage control method for coordinating based on transmission and distribution network comprehensive benefit of Fig. 1 description according to the embodiment of the invention, may further comprise the steps:

Step S110: electrical network superior control centre of subordinate uploads the critical point characteristic, and wherein, the critical point characteristic comprises regulating power, demand for control and optimization benefit.

This method is introduced three critical point characteristics of subordinate's electrical network, comprising: regulating power, demand for control and optimization benefit.

(1), regulating power.But regulating power is used for expression subordinate electrical network to the idle modification scope in critical point.

The idle Q in critical point _LiMainly be subjected to subordinate's electric network influencing, therefore need subordinate's electrical network to upload each critical point regulating power.The idle Q in critical point _LiIn, power distribution network is the system loading place, and because inside generally do not contain generator, idle control device is mainly capacity reactance device group, the first-class voltage adjusting device of on-load voltage regulation tap, so power distribution network mainly is that switching by capacity reactance device group exerts an influence to the power distribution network critical point is idle.In addition, it is idle that higher level's power transmission network also can affect subordinate's power distribution network injection, mainly contains following three aspects:

1, change in voltage causes capacitive reactance device state to change, and then it is idle to affect the injection of power distribution network critical point.

2, the acting in conjunction of change in voltage and static load characteristics causes the idle variation in distribution critical point.

3, change in voltage causes the reactive loss of distribution network to change.

When change in voltage is larger, during such as voltage collapse, 1,2 liang of aspect just has a considerable influence to the critical point is idle.In reality under normal circumstances, power transmission network change in voltage amplitude is less, so power transmission network belongs among a small circle the idle adjusting in critical point and regulate, and the critical point is idle generally to be regulated by power distribution network.

(2), demand for control.It is qualified that demand for control is used for control subordinate line voltage, and shown in the critical point voltage restriction range that needs the higher level to control.Qualified in order to guarantee subordinate's line voltage, as regulated and control by power transmission network critical point voltage U _HNeed to be controlled in certain restriction range.Therefore, need subordinate's electrical network to upload U according to current load level _HRange of needs.Aspect the voltage of critical point, power transmission network transfers the main transformer impedance much smaller in the equiva lent impedance of power distribution network with comparing, so the control of power distribution network relevant device is very little to the effect of critical point voltage, critical point voltage is affected by power transmission network mainly.

(3), optimize benefit.Optimize benefit for when control virtual controlling partition zone optimizing strategy does not clash on benefit, each subordinate's grid benefit of in the coordination optimization computation model, considering.Wherein, the virtual controlling partition zone optimizing can be thought a kind of method of the superior and the subordinate's reactive voltage coordination optimization.Each subordinate's grid net loss summation accounts for the whole network network loss ratio and sometimes cannot ignore.Do not clash in order to guarantee that the superior and the subordinate's reactive Voltage Optimum is controlled on the benefit, need in the coordination optimization computation model, introduce the optimization benefit of each subordinate's electrical network.Wherein, the coordination optimization computation model is virtual controlling partition zone optimizing recited above, and a virtual controlling subregion has comprised a part corresponding with this subordinate's electrical network in subordinate's electrical network and the higher level's electrical network.

In one embodiment of the invention, optimize benefit according to the electrical network AVC of subordinate, take the critical point variable as independent variable, by curve form or the statement of interval form.

In higher level's model, subordinate's electrical network occurs with the form of critical point variable.When electrical network superior control centre of subordinate uploaded the critical point characteristic, subordinate's electrical network changed into respectively two critical points idle Q in variable critical point with regulating power, demand for control _LiWith the critical point voltage U _HiConstraint.The expression-form of optimizing benefit can have following three kinds:

(1), perfect information model tormulation.Optimize benefit by the electrical network perfect information model f of subordinate _Li(x) accurate expression.This form is suitable for the unified situation of the superior and the subordinate's electric network information modeling.The situation that power transmission network and power transmission network were coordinated during for example net was economized and coordinated.The application benefit of subordinate's electrical network perfect information model is also comparatively remarkable, but and be not suitable for the situation that the skimble-scamble power transmission network of the superior and the subordinate's information model and power transmission network are coordinated because in this case, the many non-critical points of subordinate's electrical network information is at higher level's electrical network observable not.

(2), parameter of curve is expressed.Based on current state of electric distribution network, according to the real-time virtual mass data, will optimize benefit and carry out parameter fitting, then be expressed as the algebraic expression take the critical point variable as independent variable, as: f _Li(U _Hi, Q _Li).

Wherein, because the critical point voltage U _HiMainly be subjected to power transmission network control, power distribution network smaller to its effect, and the critical point voltage value is general in the different coordination optimization strategy changes not quite.Generally can be the critical point voltage U _HiRegard constant as.So f _Li(U _Hi, Q _Li) can be similar to and think the idle Q in critical point _LiVariable expression, shown in the formula specific as follows:

f _Li(U _Hi,Q _Li)=f _Li(Q _Li|U _Hi≈1.0)=a _inQ _Li ⁿ+a _in-1Q _Li ^n-1+...+a _i1Q _Li+a _i0

Wherein, a _Im, a _Im-1A _I0Be coefficient.The key of this expression way need to be the real-time virtual mass data.Virtual data is to provide different critical point variablees by supposition higher level control centre, under current distribution state, repeatedly calls the AVC of subordinate algorithm and calculates by regulating available optimum benefit.The distribution state can not be enumerated by total number at off-line, and needs the coefficient in the multinomial above the real-time fitting, has increased a large amount of extra calculation costs.

(3), the formal representation of piecewise function.For partly not building or building the power distribution network of AVC, adopt the expression-form of (two) described matched curve to realize that difficulty is larger, therefore, be easier in the actual conditions provide with interval form and come subsection fuzzy to express benefit.For example, for the power distribution network that still is in the reactive power compensator construction period, needs higher level control centre to support, idle power grid control can not be regarded merely the continuous control process as, so its benefit expression also should be adopted the piecewise function form.For built AVC system but the weak situation of its ability, subordinate's electrical network is difficult to increase the real-time virtual data and current benefit is carried out the function such as parameter fitting, also can adopt the statement of piecewise function form to optimize benefit.The piecewise function form is as follows:

$f_{\mathrm{Li}}\left(Q_{\mathrm{Li}}\right)=\left\{\begin{array}{cc}{f}_{i1}& Q_{\mathrm{Li}}\&Element;{\mathrm{<figure-callout\; id="1"\; label="S\; i"\; filenames="HDA00002494337900071.png,HDA00002494337900072.png"\; state="\{\{state\}\}">S</figure-callout>}}_i=[{\mathrm{<figure-callout\; id="1"\; label="l\; i"\; filenames="HDA00002494337900071.png,HDA00002494337900072.png"\; state="\{\{state\}\}">l</figure-callout>}}_i,u_{i1}]\\ f_{i2}& Q_{\mathrm{Li}}\&Element;{\mathrm{<figure-callout\; id="2"\; label="S\; i"\; filenames="HDA00002494337900021.png,HDA00002494337900071.png"\; state="\{\{state\}\}">S</figure-callout>}}_i=[{\mathrm{<figure-callout\; id="2"\; label="l\; i"\; filenames="HDA00002494337900021.png,HDA00002494337900071.png"\; state="\{\{state\}\}">l</figure-callout>}}_i,u_{i2}]\\ & ...\\ f_{\mathrm{in}}& Q_{\mathrm{Li}}\&Element;{\mathrm{<figure-callout\; id="2"\; label="S\; i"\; filenames="HDA00002494337900021.png,HDA00002494337900071.png"\; state="\{\{state\}\}">S</figure-callout>}}_i=[l_{\mathrm{in}},u_{\mathrm{in}}]\end{array}\right.,$

Wherein, S ₁∪ S ₂∪ ...=S _i, S _iBe the idle Q in critical point _LiFeas ible space.Identical with the distribution demand for control, the subinterval non-overlapping copies.Each interval benefit is got its average level usually in the following formula.The given of average level not only can be based on real time data, also can be based on Expert Rules or operating experience, and applicability is stronger.

Step S120: higher level control centre is converted into coordination variable constraint and the target correction term that characterizes with the critical point variable with the critical point characteristic, sets up reactive voltage coordination optimization world model of the superior and the subordinate.

Coordinate in the control at the superior and the subordinate's reactive voltage, coordination variable is to carry out between the superior and the subordinate mutual key, contains the information of coordinating to realize in the control global coordination optimization.The formulation of the selection of coordination variable, coordination variable constraint, coordination variable have determined the implementation effect of the superior and the subordinate's coordination controls jointly on coordinating the impact of target.

The coordination variable of this method characterizes with the critical point variable, according to the critical point characteristic of receiving, adopts the idle Q in the on high-tension side critical point of system interconnection transformer commonly used at present _LiCritical point voltage U with the power transmission network root node _HRepresent.

For example, suppose to exist 1 higher level control centre, lower N sheet net, the wherein N of connecing ₁Individual sheet network optimization benefit adopts parameter of curve to express N-N ₁Individual distribution is optimized benefit and is adopted segmentation letter segment table to reach.In conjunction with critical point recited above characteristic, consider that the reactive voltage coordination optimization world model of the superior and the subordinate of transmission and distribution network comprehensive benefit can be written as,

min f _H(x,u,Q _Li,U _Hi)+∑f _Li(Q _Li)

s.t.g _i(x,u,U _Hi,Q _Li)=0

h _j(x,u,U _Hi,Q _Li)≤0

x _min<x<x _max

u _min<u<u _max

Q _Li，min<Q _Li<Q _Li，max i＝1,2,…,N

U _Hi，min<U _Hi<U _Hi，max i＝1,2,…,N

Wherein, x, u are respectively higher level's electric network state amount and controlled quentity controlled variable, and equality constraint mainly is trend constraint, and inequality constraints mainly is the constraint of higher level's reactive-load adjusting device and critical point adjustable range and critical point demand for control that subordinate uploads, cumulative ∑ f in the target function _Li(Q _Li) be the optimization benefit of each power distribution network of subordinate, specifically be expressed as:

$f_{\mathrm{Li}}\left(Q_{\mathrm{Li}}\right)=\left\{\begin{array}{c}{y}_i=a_{\mathrm{in}}{Q_{\mathrm{Li}}}^n+a_{\mathrm{in}-1}{Q_{\mathrm{Li}}}^{n-1}+...+a_{\mathrm{<figure-callout\; id="0"\; label="i"\; filenames="HDA00002494337900031.png,HDA00002494337900032.png"\; state="\{\{state\}\}">i</figure-callout>}0},i\&Element;{\mathrm{<figure-callout\; id="1"\; label="C\; N"\; filenames="HDA00002494337900071.png,HDA00002494337900072.png"\; state="\{\{state\}\}">C</figure-callout>}}_{N_{}}\\ \\ ...\\ y_j=\left\{\begin{array}{cc}{f}_{j1}& Q_{{\mathrm{Li}}^{\&prime;}}\&Element;{\mathrm{<figure-callout\; id="0"\; label="S\; j"\; filenames="HDA00002494337900031.png,HDA00002494337900032.png"\; state="\{\{state\}\}">S</figure-callout>}}_j=[{\mathrm{<figure-callout\; id="0"\; label="l\; j"\; filenames="HDA00002494337900031.png,HDA00002494337900032.png"\; state="\{\{state\}\}">l</figure-callout>}}_j,u_{j0}]\\ f_{j2}& Q_{{\mathrm{Lj}}^{\&prime;}}\&Element;{\mathrm{<figure-callout\; id="1"\; label="S\; j"\; filenames="HDA00002494337900071.png,HDA00002494337900072.png"\; state="\{\{state\}\}">S</figure-callout>}}_j=[{\mathrm{<figure-callout\; id="1"\; label="l\; j"\; filenames="HDA00002494337900071.png,HDA00002494337900072.png"\; state="\{\{state\}\}">l</figure-callout>}}_j,u_{j1}]\\ & .\\ & .\\ & .\\ & \\ & \\ f_{\mathrm{jn}}& Q_{{\mathrm{Lj}}^{\&prime;}}\&Element;S_{\mathrm{jn}}=[l_{\mathrm{jn}},u_{\mathrm{jn}}]\end{array}\right.,j\&Element;C_{N-{\mathrm{<figure-callout\; id="1"\; label="N"\; filenames="HDA00002494337900071.png,HDA00002494337900072.png"\; state="\{\{state\}\}">N</figure-callout>}}_1}\\ ...\end{array}\right.$

Wherein, S ₁∪ S ₂∪ ...=S _i, S _iBe the idle Q in critical point _LiFeas ible space, a _Im, a _Im-1... a _I0Be coefficient.

Step S130: according to reactive voltage coordination optimization world model of the superior and the subordinate, the higher level is administered electrical network carry out the virtual controlling subregion, obtain the coordination optimization computation model of a plurality of virtual controlling subregions.

In one embodiment of the invention, higher level control centre is optimized calculating, obtains each control partition VACE set point.

Reactive voltage coordination optimization world model of the superior and the subordinate contains piecewise function, and model is complicated.The below introduces control partition, and provides Decoupled Model and Practical resolving Algorithm thereof, with further minimizing calculation scale and algorithm complex, improves computational efficiency and reliability.

VACE(Voltage/Var Area Control Error, regional REACTIVE POWER/VOLTAGE CONTROL Deviation Indices), be expressed as:

I _VACE=(Q _B-Q _Bref)+E _S(U _B-U _Bref)。

Wherein, E _S=-K _QSDistrict section network paramter matrix, irrelevant with its external network, Q _BBe critical point reactive power vector, U _BBe the critical point voltage vector.When the one's respective area does not have idle disturbance, no matter how the outside changes, and affects one's respective area I by the critical point variable _VACEValue all is zero.If the deviation of each element approximate zero all among each regional idle deviation VACE after the control, then each intra-zone reactive load changes and can ignore to outside other regional effects.Therefore, the idle VACE vector by controlling each sub regions is near null value, but the control of the decoupling zero between feasible region.Constraint below all subregion AVC control system adds:

-δ≤I _i,VACE+C _i,VACEΔu' _i≤δ

Wherein, C _{I, VACE}That control node in one's respective area is to the control sensitivity of VACE, Δ u _i' be control deviation, δ is the VACE maximum deviation that regional voltage control system allows, fibrous root factually border situation is got suitable minimum, guarantees existing feasible solution, also reaches the decoupling zero effect.Big or small comparison definition in the formula be vector element in order correspondence carry out numeric ratio.

In one embodiment of the invention, according to the Decoupling Characteristics of VACE, the reactive voltage coordination optimization world model of the superior and the subordinate that considers the transmission and distribution network comprehensive benefit is decomposed, obtain each control partition coordination optimization computation model.

Wherein, obtain each control partition coordination optimization computation model:

min f _H(x',u',Q _Li,U _Hi)+∑f _Li(Q _Li)

s.t.g _i(x',u',U _Hi,Q _Li)=0

h _j(x',u',U _Hi,Q _Li)≤0

x' _min<x'<x' _max

u' _min<u'<u' _max

Q _Li，min<Q _Li<Q _Li，max i＝1,2,…,N'

U _Hi，min<U _Hi<U _Hi，max i＝1,2,…,N'

-δ≤I _i，VACE+C _i,VACEΔu' _i≤δ

Wherein, x ', u ' are the subsets of higher level's electric network state amount x and controlled quentity controlled variable u, quantity of state and the controlled quentity controlled variable of expression control partition i inside, U _Hi, Q _LiThe critical point coordination variable of subordinate's electrical network of correspondence in the control partition i, Δ u _i' be the control deviation in the control partition i.

Step S140: higher level control centre is according to the coordination optimization computation model of virtual controlling subregion, calculate subordinate's electrical network the critical point coordination variable optimal value and be sent to subordinate's electrical network.

Calculate the optimal value of the critical point coordination variable of subordinate's electrical network.Be optimized calculating in higher level control centre, obtain the voltage U on each control partition interconnection _{IB, opt}With idle

Wherein, computation model is as follows:

min f _H(x,u,Q _Li,U _Hi)+∑f _Li(Q _Li)

s.t.g _i(x,u,U _Hi,Q _Li)=0

h _j(x,u,U _Hi,Q _Li)≤0

x _min<x<x _max

u _min<u<u _max

For control partition i, with current critical point voltage vector on the outside link line be U _IB, current critical point reactive power vector is Q _IB, regional VACE coefficient matrix is E _IS, the VACE vector of controlled subregion i.Set up the coordination optimization computation model that correspondence is permitted in this control minute at each control partition.Wherein, if if there is not piecewise function in target function in the control partition coordination optimization computation model, namely not with piecewise function statement subordinate grid benefit characteristic, then directly obtain higher level's power grid control amount and the critical point coordination variable that instructs subordinate's optimisation strategy by control partition coordination optimization computation model.

If there is piecewise function in target function in the control partition coordination optimization computation model, then just need to upload benefit before subregion is that the sheet net of segmentation form is processed, and marks separately a subregion and its coordination.For the latter, can be by enumerating the optimum benefit in each feasible subspace, obtain corresponding computation model as shown in the formula:

min f _H(x',u',Q _Li,U _Hi)+f _ij

s.t.g _i(x',u',U _Hi,Q _Li)=0

h _j(x',u',U _Hi,Q _Li)≤0

x' _min<x'<x' _max

u' _min<u'<u' _max

Q _Lij∈S _ij=[l _ij,u _ij] i=1,2,…,N,1≤j≤n

U _Hi，min<U _Hi<U _Hi,max i＝1,2,…,N'

-δ≤I _i,VACE+C _i,VACEΔu' _i≤δ

Wherein, Q _{Li, min}＜Q _Li＜Q _{Li, max}Be set to the subspace Q of critical point variable _Lij∈ S _Ij=[l _Ij, u _Ij], target function f _Li(Q _Li) become constant f from piecewise function _Ij, can be in the hope of the optimal solution x of this control partition i at constraint subspace j _{Ij, opt}, and corresponding optimum network loss f _{HL, ij, opt}More all retrain at last the optimal solution f of subspace _{HL, i, opt}, can obtain the optimal solution in whole coordination subarea, and then obtain considering the coordination subarea optimisation strategy of the superior and the subordinate's electric network synthetic benefit, and each critical point coordination variable optimal value of subordinate's power distribution network corresponding to control partition.

Higher level control centre will calculate subordinate's electrical network the critical point coordination variable optimal value and be sent to subordinate's electrical network.

In one embodiment of the invention, also comprise the improvement algorithm, be used for the virtual controlling subregion is optimized calculating that integrate and generate new virtual partition, wherein, the number of new virtual partition is less than the virtual partition number before optimizing.

When the interval number of optimizing benefit is more, need corresponding Optimal calculation trend calculation times more, for example contain 10 intervals, then need 10 suboptimum Load Flows to calculate.Therefore, according to the characteristics of feasible constraint space and piecewise function, propose to improve derivation algorithm, can effectively reduce control partition Optimal calculation trend calculation times.The interval non-overlapping copies of general benefit expression form can be summarized as follows formula:

$f_{\mathrm{Li}}\left(Q_{\mathrm{Li}}\right)=\left\{\begin{array}{cc}{f}_{i1}& Q_{\mathrm{Li}}\&Element;{\mathrm{<figure-callout\; id="1"\; label="S\; i"\; filenames="HDA00002494337900071.png,HDA00002494337900072.png"\; state="\{\{state\}\}">S</figure-callout>}}_i=[{\mathrm{<figure-callout\; id="1"\; label="l\; i"\; filenames="HDA00002494337900071.png,HDA00002494337900072.png"\; state="\{\{state\}\}">l</figure-callout>}}_i,u_{i1}]\\ f_{i2}& Q_{\mathrm{Li}}\&Element;{\mathrm{<figure-callout\; id="2"\; label="S\; i"\; filenames="HDA00002494337900021.png,HDA00002494337900071.png"\; state="\{\{state\}\}">S</figure-callout>}}_i=[{\mathrm{<figure-callout\; id="2"\; label="l\; i"\; filenames="HDA00002494337900021.png,HDA00002494337900071.png"\; state="\{\{state\}\}">l</figure-callout>}}_i,u_{i2}]\\ & ...\\ f_{\mathrm{in}}& Q_{\mathrm{Li}}\&Element;{\mathrm{<figure-callout\; id="2"\; label="S\; i"\; filenames="HDA00002494337900021.png,HDA00002494337900071.png"\; state="\{\{state\}\}">S</figure-callout>}}_i=[l_{\mathrm{in}},u_{\mathrm{in}}]\end{array}\right.$

Wherein, f _IjFor being i the network loss level of coordinating j constraint subspace, subarea, S ₁∪ S ₂∪ ...=S _i, S _iBe the idle Q in critical point _LiFeas ible space, and

1＜=j, k＜=n.General commonly used network loss is as performance indicator, and it is virtual that power distribution network relation between the idle and minimum network loss in critical point when the variable of given critical point is carried out mass data, and can obtain both is the relation of single paddy function in constraint space.

In constraint space, according to the horizontal f of the network loss of each subspace _Ij, resequenced in the interval, and adjacent interval is merged, obtain the following interval modification of expressing benefit of optimizing:

Further each interval of constraint space is changed into the form of the modification of optimizing interval expression benefit, the meaning of interval representative becomes: the more excellent benefit that can reach in interval.Along with the increase of interval rank numbering, progressively relax in the interval, and the network loss level descends.

According to revised benefit statement, computation model below in this control partition, adopting:

min f _H(x',u',Q _Li,U _Hi)

s.t.g _i(x',u',U _Hi,Q _Li)=0

h _j(x',u',U _Hi,Q _Li)≤0

x' _min<x'<x' _max

u' _min<u'<u' _max

Q _Lij∈S' _ij=[l' _ij,u' _ij] i=1,2,…,N,1≤j≤n

U _Hi，min<U _Hi<U _Hi，max i＝1,2,…,N'

-δ≤I _i,VACE+C _i,VACEΔu' _i≤δ

Improvement algorithm for the interval statement comprises the steps:

Step S141: select constraint subspace (being full constraint space) S _In' begin to calculate.

Step S142: the constraint subspace that makes current calculating is S _Ij', ask following computation model:

min f _H(x',u',Q _Li,U _Hi)

s.t.g _i(x',u',U _Hi,Q _Li)=0

h _j(x',u',U _Hi,Q _Li)≤0

x' _min<x'<x' _max

u' _min<u'<u' _max

Q _Lij∈S' _ij=[l' _ij,u' _ij] i=1,2,…,N,1≤j≤n

U _Hi，min<U _Hi<U _Hi，max i＝1,2,…,N'

-δ≤I _i,VACE+C _i,VACEΔu' _i≤δ

Wherein, x ', u ' are the subset of higher level's electric network state amount x and controlled quentity controlled variable u, and U is critical point voltage, and Q is that the critical point is idle, and min and max represent respectively corresponding minimum and maximum threshold value, I _{I, VACE}Be the VACE index of constraint subspace i, C _{I, VACE}That constraint subspace i is to the control sensitivity of VACE, Δ u _i' be control deviation, δ is the VACE maximum deviation that regional voltage control system allows.Calculate locally optimal solution f _{Ij, opt}', can obtain coordinating subarea i constraint subspace S _In' locally optimal solution drop point x _Ij' and locally optimal solution f _{HL, ij, opt}, wherein,

f _HL,ij,opt=f' _ij,opt+f' _ij

Step S143: if drop point belongs to other subspaces S _Ik', basis then

Relation, obtain subspace S _Ik' locally optimal solution f _{HL, ik, opt}

f _H,ik，opt=f' _ij,opt+f' _ik

If drop point x _Ij' belonging to a plurality of spaces, the interval numbering that interval rank numbering is minimum is designated as N _TEMPBecause

F ' _Ik＜f' _Ij, interval rank numbering N _TEMPLocally optimal solution f _{HL, ik, opt}Be the relatively minimum solution of the network loss in above-mentioned a plurality of spaces.

Step S144: if calculate for the first time, remember interval rank numbering N _TEMPLocally optimal solution f _{HL, ik, opt}Be f _{HL, i, temp}, and remember current drop point x _Ij' be x _{I, temp}Once calculated more current f if carried out _{HL, ik, opt}F with record _{HL, i, temp}If satisfy f _{HL, ik, opt}＜f _{HL, i, temp}, then make f _{HL, i, temp}=f _{HL, ik, opt}, x _{I, temp}=x _Ij'; Otherwise, if satisfy f _{HL, i, temp}＜f' _{Ij, opt}+ f _I1', then improve algorithm iteration and finish.

Step S145: if interval numbering N _TEMP+ 1 is non-vanishing, gets corresponding constraint subspace

Execution in step S2.

If interval numbering N _TEMP+ 1 is zero, then improves algorithm iteration and finishes.

Step S146: single control partition i calculates and finishes the optimal solution f of controlled minute trivial i _{HL, i, opt}With optimal solution drop point x _i, wherein, f _{HL, i, opt}=f _{HL, i, temp}, x _i=x _{I, temp}

Above-mentioned improvement algorithm has designed the Practical resolving Algorithm according to the characteristics of feasible constraint space and piecewise function, and the interior Optimal calculation trend calculation times of subregion is reduced.

It should be noted that it is the hypothesis of single paddy trend at feas ible space that the interval expression formula of top benefit is based on general distribution accent benefit.When being not single paddy, can be by the new drop point of judging that minute last level interval of paddy calculates, finish the iteration of carrying out again other paddy after the iteration of this drop point place paddy, then the result is compared.

Step S150: subordinate's electrical network adds the coordination variable constraint according to the optimal value of critical point coordination variable in the AVC algorithm, obtain the virtual controlling partition zone optimizing strategy based on the transmission and distribution network comprehensive benefit.With each critical point coordination variable set point U _{Hi, ref}And Q _{Li, ref}Be handed down to each control centre of corresponding subordinate, All Control Center is used the critical point voltage U _{Hi, ref}, and adding the tracking idle constraint in critical point in the AVC algorithm separately, thus the virtual controlling partition zone optimizing strategy based on the transmission and distribution network comprehensive benefit further obtained.

Step S160: higher level's electrical network and subordinate's electrical network carry out reactive voltage according to virtual controlling partition zone optimizing strategy and coordinate control.

Below by a simulation example feasibility of the present invention and advantage are further specified, are understandable that, described simulation example only for illustrative purposes, embodiments of the invention are not limited to this.

This example adopts the IEEE39 node system to carry out simulation calculation.At first verified and introduced the feasibility that the VACE index realizes decoupling zero control between control partition, and the result of global optimization and coordination optimization compared, when then having verified respectively the access of single power distribution network and during the access of a plurality of power distribution network, the feasibility of the main website coordination optimization control strategy that the present invention proposes, with traditional power distribution network critical point ability of not considering, algorithm and the algorithm of the present invention of demand and optimization benefit compare, consider at last a plurality of power distribution network accesses, when load or burden without work increased continuously, the implementation effect of the algorithm that traditional algorithm and the present invention are proposed compared.

Realize the decoupling zero of higher level's power grid control subregion based on the VACE index.Adopt the IEEE39 node system to carry out simulation calculation.Be preset with 17 interconnections, at 1-2,3-4 cuts on three interconnections of 16-17, obtains two subregions of two IEEE39 systems, contains node { 2,3,17,18,25,26,27,28,29,30,37,38} is designated as subsystem A, contains node { Isosorbide-5-Nitrae, 5,6,7,8,9,10,11,12,13,14,15,16,19,20,21,22,23,24,31,32,33,34,35,36,39} is subsystem B, as shown in Figure 2.

Be optimized calculating at total system, the result that can be optimized, P _Loss=0.27288; Two systems carry out respectively independent optimization, obtain separately optimum results: P _ALoss=0.11937, P _BLoss=0.15438.At this moment, subsystem network loss summation is P _Loss'=P _ALoss+ P _BLoss=0.27375

As seen, the optimum results that obtains before and after the subregion is close, can think that global optimization is consistent with the result of coordination optimization.If adopt other cutting mode, for example with { 1,3,9} is included into subsystem A, also can obtain similar results.The subregion that proof realizes based on VACE does not affect the control effect of optimisation strategy.

Coordination optimization control strategy when single power distribution network access.The feasibility of main website coordination optimization control strategy during the single power distribution network access of the below's checking.Choosing five kinds of algorithms compares: the algorithm of not considering power distribution network critical point ability, demand and optimization benefit; Only consider the algorithm of power distribution network critical point ability and demand; Consider power distribution network critical point ability and demand, take curve statement benefit as algorithm; Consider power distribution network critical point ability and demand, take interval statement benefit as algorithm; Take interval statement benefit as the improvement algorithm.Wherein, the algorithm of the present invention's proposition is rear two kinds of algorithms.

Take the IEEE39 node system as example.No. 17 nodes are coordinated the critical point as transmission and distribution network, namely the power transmission network root node at No. 17 nodes to the power distribution network equivalence of loading.Coordination variable selects power transmission network root node voltage U _HInject idle Q with the power distribution network critical point _L, being assumed to by beginning regulating power and demand for control of the power distribution network that connects:

1.00<U _H<1.07

-8<Q _L<-1

Algorithm effect makes the optimum benefit of power distribution network be about half of the optimum benefit of power transmission network for convenience of description.Grid benefit is assessed by network loss, is made as

f _Distr=0.1475-0.01Q _L+0.01Q _L ²

In order mutually to verify result of calculation, the function of explaining benefit take interval form is that the section of following formula is processed, i.e. following formula:

$f_{\mathrm{Distr}}=\left\{\begin{array}{cc}0.785,& Q_L\&Element;(-8,-7]\\ 0.635,& Q_L\&Element;(-7,-6]\\ 0.505,& Q_L\&Element;(-6,-5]\\ 0.395,& Q_L\&Element;(-5,-4]\\ 0.305,& Q_L\&Element;(-4,-3]\\ 0.235,& Q_L\&Element;(-3,-2]\\ 0.185,& Q_L\&Element;(-2,-1]\end{array}\right.$

That note is not transferred critical point ability, demand and optimization benefit with considering is algorithm A ₀That only considers power distribution network critical point ability and demand is designated as algorithm A ₁Consider power distribution network critical point ability and demand, be designated as algorithm A with curve statement benefit form ₂Be designated as algorithm A with interval statement benefit form ₃Improvement algorithm with interval statement benefit form is designated as A ₄Result of calculation is seen IEEE39 node system transmission and distribution network coordination example table 1.

Table 1

As seen from Table 1, algorithm A ₀Although the result who obtains is best, because of demand and the constraint of not considering subordinate's electrical network, critical point voltage is got over the receivable upper limit of subordinate's electrical network, will cause subordinate's power distribution network key node voltage out-of-limit if implement this optimisation strategy.Algorithm A ₃, A ₄In distribution network loss only represent the level of each interval attainable optimization network loss, only be used for calculating the power transmission network optimisation strategy, obtain the critical point optimization solution, actual distribution network loss still need be decided on the idle value of issuing in critical point.Algorithm A ₁, A ₂, A ₃, A ₄The result is close, and table 1 fails to show its nuance because keeping result length.Because the expression function of subordinate's grid net loss in its constraint space without the minimum point Q of its idle coordination constraint in critical point _L=0.5, the below makes amendment to demand for control and regulating power, makes constraint space toward subordinate's grid net loss minimum point direction continuation.

$f_{\mathrm{Distr}}=\left\{\begin{array}{cc}0.785,& Q_L\&Element;(-8,-7]\\ 0.635,& Q_L\&Element;(-7,-6]\\ 0.505,& Q_L\&Element;(-6,-5]\\ 0.395,& Q_L\&Element;(-5,-4]\\ 0.305,& Q_L\&Element;(-4,-3]\\ 0.235,& Q_L\&Element;(-3,-2]\\ 0.185,& Q_L\&Element;(-2,-1]\\ 0.155,& Q_L\&Element;(-\mathrm{1,0}]\\ 0.145,& Q_L\&Element;(\mathrm{0,1}]\\ 0.155,& Q_L\&Element;\left(\mathrm{1,2}\right]\end{array}\right.$

1.00<U _H<1.07

Wherein,

-8<Q _L<2

The below recomputates according to above-mentioned modification, obtains the table 2 that IEEE39 node system transmission and distribution network is coordinated example:

Table 2

After constraint space comprised subordinate's grid net loss minimum point, contrast table 2 can be found, only considers the algorithm A of the algorithm of subordinate's ability and demand ₁The result change has occured.This algorithm is further considered the algorithm A of subordinate's benefit at that point with other ₂, A ₃, A ₄Compare, although Losses is less, subordinate's distribution network loss is higher, and the result is that the total network loss of total system increases to some extent, has reflected the implicit the superior and the subordinate's optimisation strategy conflict of such algorithm.Algorithm A ₂, A ₃, A ₄Consider subordinate's power distribution network benefit, although therefore Losses increases to some extent, the whole network loss of total system is optimized.

Regulating whether comprise in twice calculating in the minimum front and back of subordinate's grid net loss contrast algorithm A ₃With algorithm A ₄, the optimum results before and after visible discrete model is revised is basically identical.For example, A ₃Losses be 0.281296, A ₄Be 0.281298.But variant aspect computational speed, for example in twice example, A ₃Need to calculate all intervals, as in the 1st example, there being 7, have 10 in the 2nd example, and A ₄Only need to calculate 2 and 3.

Coordination optimization control strategy when a plurality of power distribution network access.During a plurality of power distribution network access of the below's checking, the feasibility of main website coordination optimization control strategy proposed by the invention.Having chosen three kinds of algorithms in the sample calculation analysis compares: the algorithm A that does not consider power distribution network critical point ability, demand and optimization benefit ₀', wherein, A ₀' be algorithms most in use; Only consider the algorithm A of power distribution network critical point ability and demand ₁'; Consider the algorithm A of power distribution network critical point ability, demand and benefit ₂', i.e. method proposed by the invention.Wherein, algorithm A ₂' be to calculate each control partition VACE set point at total system, then carry out control partition optimization and calculate.Each control partition forms corresponding control constraint, forms coordination constraint and target correction term based on the critical point characteristic based on this VACE set point, thereby sets up control partition coordination optimization computation model, calculates and obtain each subordinate's power distribution network coordination variable value of issuing.

The below is split as two subsystems A and B as shown in Figure 2 take the IEEE39 node system as example.Consider that simple two power distribution networks are linked into the situation of power transmission network, wherein No. 1 power distribution network is connected to the critical point No. 17, and power distribution network AVC ability is higher, can provide demand for control and regulating power, also can explain its benefit with curve.No. 2 power distribution network is connected to the critical point No. 13, and No. 2 power distribution network AVC ability is slightly not enough, can provide demand for control and regulating power, explain its benefit with the interval.Subsystem A and No. 1 power distribution network composition control subregion A1, subsystem B and No. 2 power distribution network composition control subregion B2.And the note do not comprise that the total system of subordinate's power distribution network is S, the total system that comprises subordinate's power distribution network is S'.By calculating IEEE39 node system transmission and distribution network tuning algorithm contrast table 3.

Table 3

In order to make three kinds of algorithm network loss possess comparativity, make the computational methods of each distribution network loss consistent, will respectively hang oneself optimize to calculate try to achieve the critical point inject idle after, further substitution f _Distr=0.1475-0.01Q _L+ 0.01Q _L ²Calculate.

As can be seen from Table 3, algorithm A ₀' owing to do not consider subordinate's power distribution network situation, although Losses is optimum, whole total system network loss is larger, and there is the out-of-limit situation of node voltage in subordinate's electrical network; Algorithm A ₁' considered control ability and the demand for control of subordinate's electrical network can guarantee the total system voltage security, but owing to reckon without subordinate's grid net loss, so whole network loss is larger; Algorithm A ₂' introduced the consideration to subordinate's grid net loss, although Losses is a little more than algorithm A ₁', but have a clear superiority in whole total system S ' network loss.

Each algorithm implementation effect when load or burden without work increases continuously.The below considers a plurality of power distribution networks accesses, when load or burden without work increases continuously, and the implementation effect of traditional algorithm and algorithm proposed by the invention relatively.Sample calculation analysis comprises the algorithm of not considering power distribution network critical point ability, demand and optimization benefit; Only consider the algorithm of power distribution network critical point ability and demand; Consider the algorithm of power distribution network critical point ability, demand and benefit.

Take the IEEE39 node system as example, consider that simple two power distribution networks are linked into the power transmission network situation, wherein No. 1 power distribution network is connected to the critical point No. 18, and No. 2 power distribution network is connected to the critical point No. 13.Algorithm A ₀', A ₁', A ₂' implication is the same; No. 1, No. 2 power distribution network critical point characteristic is also same as above.Make system from the 2nd control cycle, namely during t=2, load bus load or burden without work is by 1% sustainable growth, and the below is idle to transmission and distribution network network loss, the critical point of each algorithm in 14 control cycles, critical point voltage is analyzed.

Can be seen by Fig. 3, fall the damage angle from total system, algorithm A ₂' good than other algorithms.As shown in Figure 4, fall in the power transmission network side and decrease aspect the effect angle algorithm A ₂' slightly poor than other algorithms.Reason can be learnt algorithm A by the total network loss contrast of Fig. 5 power distribution network ₂' considered the control of power distribution network benefit, coordination optimization after, the network loss that connects power distribution network under No. 13 critical points and No. 18 critical points is all optimized.Figure Fig. 6 is to the network loss situation that gets access to grid under Fig. 9 has represented power transmission network side subsystem A, B and No. 13, No. 18.Figure 10 and Figure 11 have shown the contrast of two power distribution network critical point voltages.As we can see from the figure, algorithm A ₀' owing to there is not to consider the lower voltage requirements that connects power distribution network, critical point voltage only is controlled at 0.9 to 1.1 by general node processing, it is out-of-limit to connect the power distribution network builtin voltage under may causing.Algorithm A ₁', A ₂' then according under connect the distribution network voltage demand, well voltage control in 1.0 to 1.07 scopes.Figure 12 and Figure 13 have shown the idle contrast of two power distribution network critical points injection.As we can see from the figure, algorithm A ₀' owing to do not consider the lower idle regulating power that connects power distribution network, only think that it is adjustable, under exceeding, the 13rd, 14 control cycle Optimal Setting values connect the idle adjusting lower limit of power distribution network; Algorithm A ₁', A ₂' connect the power distribution network critical point under having considered and inject idle regulating power, the optimization that issues is calculated set point control in the reasonable scope.

To sum up, the algorithm that the present invention proposes can effectively improve system losses, and considered demand for control and the idle regulating power of subordinate to critical point voltage, thus guarantee transmission and distribution network both sides reactive balance, avoid because of the critical point voltage power distribution network internal node that causes inclined to one side or on the low side out-of-limit.Owing to adopted zoning concept, Divergent or optimum results are unavailable as a result when a certain coordination subarea optimal load flow, are to have influence on calculating and even the control that other coordinate subareas; Transfer AVC system or the algorithm that does not have or only have insufficiency of function in a plurality of coordinations subarea, when needing to adopt piecewise function to express the network loss level, the total algorithm performance is declined to a great extent, namely algorithm has certain compatibility.

The reactive voltage control method for coordinating based on the transmission and distribution network comprehensive benefit according to the embodiment of the invention has been considered subordinate's operation of power networks demand, ability and benefit, is showing the running status of having improved the whole network, has improved the system benefit of the whole network;

The algorithm that the present invention proposes has improved the system benefit of the whole network owing to improved the running status of the whole network; In the algorithm higher level's power transmission network is carried out control partition, and computation model and the Practical resolving Algorithm of proposition control partition, with the superior and the subordinate's electric network reactive-load voltage coordination problem in the large electric network reactive-load voltage Harmonic Control decomposition and inversion control partition that is dimensionality reduction, reduce computation model scale and computation complexity; Simultaneously subregion decoupling zero control has improved the reliability of AVC algorithm, and unusual if certain subregion occurs, for example optimal load flow is dispersed or optimum results is unavailable, also can not have influence on the normal coordination optimization control of other subregions.Take into full account in addition multi-form that subordinate's benefit uploads in the algorithm, take into full account subordinate and upload and optimize the multi-form of benefit, improved the compatibility of the benefit expression-form that traditional algorithm uploads subordinate, reduced the extra calculation cost of subordinate electrical network.

In the description of this specification, the description of reference term " embodiment ", " some embodiment ", " example ", " concrete example " or " some examples " etc. means to be contained at least one embodiment of the present invention or the example in conjunction with specific features, structure, material or the characteristics of this embodiment or example description.In this manual, the schematic statement of above-mentioned term not necessarily referred to identical embodiment or example.And the specific features of description, structure, material or characteristics can be with suitable mode combinations in any one or more embodiment or example.

Although illustrated and described embodiments of the invention, for the ordinary skill in the art, be appreciated that without departing from the principles and spirit of the present invention and can carry out multiple variation, modification, replacement and modification to these embodiment that scope of the present invention is by claims and be equal to and limit.

Claims (8)

1. the reactive voltage control method for coordinating based on the transmission and distribution network comprehensive benefit is characterized in that, may further comprise the steps:

Electrical network superior control centre of subordinate uploads the critical point characteristic, and wherein, described critical point characteristic comprises regulating power, demand for control and optimization benefit;

Described higher level control centre is converted into coordination variable constraint and the target correction term that characterizes with the critical point variable with described critical point characteristic, sets up reactive voltage coordination optimization world model of the superior and the subordinate;

According to reactive voltage coordination optimization world model of described the superior and the subordinate, described higher level is administered electrical network carry out the virtual controlling subregion, obtain the coordination optimization computation model of a plurality of virtual controlling subregions;

Described higher level control centre is according to described coordination optimization computation model, calculate described subordinate electrical network the critical point coordination variable optimal value and be sent to described subordinate electrical network;

Described subordinate electrical network adds the coordination variable constraint according to the optimal value of described critical point coordination variable in automatism voltage control AVC algorithm, obtain the virtual controlling partition zone optimizing strategy based on the transmission and distribution network comprehensive benefit; And

Described higher level's electrical network and described subordinate electrical network carry out reactive voltage according to described virtual controlling partition zone optimizing strategy and coordinate control.

2. the reactive voltage control method for coordinating based on the transmission and distribution network comprehensive benefit as claimed in claim 1 is characterized in that, but described regulating power is used for representing that described subordinate electrical network is to the idle modification scope in described critical point.

3. the reactive voltage control method for coordinating based on the transmission and distribution network comprehensive benefit as claimed in claim 1 is characterized in that, it is qualified that described demand for control is used for controlling described subordinate line voltage, and shown in the critical point voltage restriction range that needs the higher level to control.

4. the reactive voltage control method for coordinating based on the transmission and distribution network comprehensive benefit as claimed in claim 1, it is characterized in that, described optimization benefit is when the described virtual controlling partition zone optimizing strategy of control does not clash on benefit, each subordinate's grid benefit of considering in the coordination optimization computation model.

5. as claimed in claim 1 based on the reactive voltage control method for coordinating of transmission and distribution network comprehensive benefit, it is characterized in that described optimization benefit is according to the electrical network AVC of subordinate, take described critical point variable as independent variable, by curve form or the statement of interval form.

6. as claimed in claim 1 based on the reactive voltage control method for coordinating of transmission and distribution network comprehensive benefit, it is characterized in that, also comprise the improvement algorithm, be used for the virtual controlling subregion is optimized calculating, integrate and generate new virtual partition, wherein, the number of described new virtual partition is less than the virtual partition number before optimizing.

7. the reactive voltage control method for coordinating based on the transmission and distribution network comprehensive benefit as claimed in claim 1 is characterized in that, described higher level control centre is optimized calculating, obtains each control partition VACE set point.

8. as claimed in claim 6 based on the reactive voltage control method for coordinating of transmission and distribution network comprehensive benefit, it is characterized in that, described the virtual controlling subregion be optimized calculating, integrate and generate new virtual partition and further comprise:

Step 1: select constraint subspace S _In' begin to calculate;

Step 2: making the described constraint subspace of current calculating is S _Ij', find the solution following computation model:

min f _H(x',u',Q _Li,U _Hi)

s.t.g _i(x',u',U _Hi,Q _Li)=0

h _j(x',u',U _Hi,Q _Li)≤0

x' _min<x'<x' _max

u' _min<u'<u' _max

Q _Lij∈S' _ij=[l' _ij,u' _ij]i=1,2,…,N,1≤j≤n

U _Hi，min<U _Hi<U _Hi，max i＝1,2,…,N'

-δ≤I _i,VACE+C _i,VACEΔu' _i≤δ

Wherein, x ', u ' are the subset of higher level's electric network state amount x and controlled quentity controlled variable u, and U is described critical point voltage, and Q is that described critical point is idle, and min and max represent respectively corresponding minimum and maximum threshold value, I _{I, VACE}Be the VACE index of described constraint subspace i, C _{I, VACE}Be the control sensitivity of described constraint subspace i to VACE, Δ u _i' be control deviation, δ is the VACE maximum deviation that regional voltage control system allows.Calculate locally optimal solution f _{Ij, opt}', can obtain coordinating subarea i constraint subspace S _In' locally optimal solution drop point x _Ij' and locally optimal solution f _{HL, ij, opt}, wherein,

f _{HL，ij，opt}=f' _ij，opt+f' _ij

Step 3: if drop point x _Ij' belong to other subspaces S _Ik', basis then

Relation, obtain subspace S _Ik' locally optimal solution f _{HL, ik, opt}, wherein, f _{HL, ik, opt}=f' _{Ij, opt}+ f ' _Ik

If described drop point x _Ij' belong to a plurality of subspaces, then the interval numbering that interval rank numbering is minimum is designated as N _TEMPBecause

F ' _Ik＜f' _Ij, described interval rank numbering N _TEMPLocally optimal solution f _{HL, ik, opt}Be the relatively minimum solution of the network loss in described a plurality of spaces.

Step 4: if calculate for the first time, remember described interval rank numbering N _TEMPDescribed locally optimal solution f _{HL, ik, opt}Be f _{HL, i, temp}, and remember current drop point x _Ij' be x _{I, temp}Once calculated more described f if carried out _{HL, ik, opt}With described f _{HL, i, temp}If satisfy f _{HL, ik, opt}＜f _{HL, i, temp}, then make f _{HL, i, temp}=f _{HL, ik, opt}, x _{I, temp}=x _Ij'; Otherwise, if satisfy f _{HL, i, temp}＜f' _{Ij, opt}+ f _I1', then improve algorithm iteration and finish.

Step 5: if described interval numbering N _TEMP+ 1 is non-vanishing, gets corresponding described constraint subspace

Execution in step 2.

If described interval numbering N _TEMP+ 1 is zero, then improves algorithm iteration and finishes.

Step 6: single described control partition i calculates and finishes, and obtains the described optimal solution f of described control partition i _{HL, i, opt}With optimal solution drop point x _i, wherein, f _{HL, i, opt}=f _{HL, i, temp}, x _i=x _{I, temp}

Images