CN104617585A - Reactive compensation configuration method - Google Patents

Abstract

The invention discloses a reactive compensation configuration method, in particular relates to an active optimizing method of a generation and transmission grid including probability fluctuation power connection, and belongs to the technical field of reactive optimization of power systems. The method comprises the steps of forming a node admittance matrix in the power system, and setting initial voltage of each node; selecting control variable and state variable; building a reactive optimization change limitation mathematical model which specially includes target functions, portability characteristic limiting conditions and fluctuation characteristics limitation conditions; determining the configuration of the control variable by the stochastic particle swarm optimization based on the latin hypercube probabilistic power flow. With the adoption of the method, a reactive compensation device in the transmission grid can be effectively configured according to the power probability characteristics and fluctuation characteristics of the grid connection of a large-scale photovoltaic station, and therefore, the grid can run normally based on the probability meanings, and moreover, the voltage fluctuation of the photovoltaic power can be greatly reduced, and as a result, the voltage and electricity energy quality of the grid can be increased.

Description

Reactive compensation configuration method

Technical field

The invention discloses reactive compensation configuration method, particularly relate to the power transmission network idle work optimization method containing probability fluctuation plant-grid connection, belong to the technical field of reactive power optimization of power system.

Background technology

Along with the consumption day by day of fossil energy, traditional energy cannot meet the demand of human social development to energy resource supply, regenerative resource more and more accept by the mankind and develop.Solar energy is as inexhaustible clear energy sources, and receive profound research and concern, particularly traditional solar energy-heat energy technology is widely applied in every respect.Electric energy all has the energy mode of high efficiency and flexibility as utilization ratio in energy resource system and type of service, has been the requisite daily necessities of huge numbers of families.In conjunction with the demand of society and the development of science and technology, solar grid-connected generation technology has become the problem of global scientific and technological circle and industrial quarters, and obtains effective investigation and application.

The conversion process of solar energy-electric energy, i.e. photovoltaic effect, be that solar energy is converted into the process of electric energy by photovoltaic battery panel, be also called photovoltaic generation.There is all many-sides in the challenge of photovoltaic power generation grid-connecting technology, from device aspect, the technique of photovoltaic battery panel, the manufacture etc. of combining inverter are all research emphasis; From electrical network aspect, the electric energy management after grid-connected, active power dispatch and reactive power compensation are all that operation of power networks brings new problem.Particularly the probability and fluctuation of grid-connected rear power input, directly results in the difficulty of power transmission network reactive power compensation.

Traditional idle planning problem, refer to by configuring suitable reactive-load compensation equipment, by adjusting generator voltage in operation of power networks, the mode such as transformer divides tap, reactive-load compensation equipment power stage, realize electrical network under normal operating conditions, every voltage indexes meets the demands, and reduces network transmitting power loss as far as possible.Therefore traditional idle planning has been the optimization problem of a multivariable, multiple constraint, non-linear, discrete variable-continuous variable mixing.After adding photovoltaic generation, the research for photovoltaic random chance has been applied to reactive power compensation planning problem, but after pv grid-connected plant reaches certain scale, the fluctuation of photovoltaic plant cluster power stage is remarkable.Therefore, for the reactive power compensation allocation problem of large-scale photovoltaic energy access, not only to consider that photovoltaic power exports probability, more will consider photovoltaic power wave characteristic in time, to realize effective reactive power compensation.

Summary of the invention

Technical problem to be solved by this invention is the deficiency for above-mentioned background technology; provide reactive compensation configuration method; can for the power probability characteristic of large-scale photovoltaic electric station grid connection and wave characteristic; realize effective configuration of power transmission network reactive-load compensation equipment; the normal operation of electrical network is ensure that in operation problem not only on probability meaning; and the voltage fluctuation that the fluctuation greatly reducing photovoltaic power causes, achieve the lifting of line voltage and the quality of power supply.

The present invention adopts following technical scheme for achieving the above object:

Reactive compensation configuration method, adopts shunt capacitor and static reacance generator to carry out collaborative configuration compensation, specifically comprises the steps:

(1) in electric power system, form node admittance matrix and set each node initial voltage;

(2) control variables and state variable is chosen;

(3) set up idle work optimization chance constraint Mathematical Modeling, specifically comprise target function, probability characteristics constraints and fluctuation characteristic constraints, wherein, fluctuation constraints:

$\left\{\begin{array}{c}{P}_i+\mathrm{\&Delta;}{P}_i-{V_i}^{\&prime;}\underset{j\&Element;i}{\mathrm{\&Sigma;}}{V_j}^{\&prime;}(G_{\mathrm{ij}}\cos {{\mathrm{\&theta;}}_{\mathrm{ij}}}^{\&prime;}+B_{\mathrm{ij}}\sin {{\mathrm{\&theta;}}_{\mathrm{ij}}}^{\&prime;})=0\\ Q_i+\mathrm{\&Delta;}{Q}_{{\mathrm{SVG}}_i}-{V_i}^{\&prime;}\underset{j\&Element;i}{\mathrm{\&Sigma;}}{V_j}^{\&prime;}(G_{\mathrm{ij}}\sin {{\mathrm{\&theta;}}_{\mathrm{ij}}}^{\&prime;}-B_{\mathrm{ij}}\cos {{\mathrm{\&theta;}}_{\mathrm{ij}}}^{\&prime;})=0\\ P\{\mathrm{\&Delta;}{P_i}^{-}\&le;\mathrm{\&Delta;}{P}_i\&le;\mathrm{\&Delta;}{P_i}^{+},i=\mathrm{1,2},...,n\}>\mathrm{\&beta;}\\ \mathrm{abs}(V_i-{V_i}^{\&prime;})<\mathrm{Req},i=\mathrm{1,2},...,n\end{array}\right.,$ Wherein,

P _ifor the active power that node i place is injected, Δ P _ifor the variable quantity of active power is injected at node i place, V _ifor the voltage magnitude at node i place, V _i' be the voltage magnitude at node i place during power fluctuation, V _j' be the voltage magnitude at node j place during power fluctuation, G _ijfor network conductance, B _ijfor network susceptance, θ _ij' be the voltage phase difference of node i during power fluctuation and node j, Q _ifor the reactive power that node i place is injected, for during power fluctuation, node i place reacance generator sends the variable quantity of reactive power, β is the confidence bound level for wave characteristic, Δ P _i ^-for the minimum value in confidence bound level lower node i place active power fluctuation, Δ P _i ⁺for the maximum in confidence bound level lower node i place active power fluctuation, n is interstitial content, and abs () is signed magnitude arithmetic(al), and Req is the scope of admissible voltage fluctuation.

(4) configuration based on the random particles group algorithm determination control variables of Latin hypercube Probabilistic Load Flow is adopted.

As the further prioritization scheme of described reactive compensation configuration method, choose generator reactive power in step (2), shunt capacitor capacity, adjustable transformer divide tap to be control variables, active power, the node voltage of access line transmission are state variable.

As the further prioritization scheme of described reactive compensation configuration method, in step (3), set up target function f, minf=μ Q so that system equivalence annual operating cost is minimum for target _cost+ P _cost, Q _costfor reactive power compensation configuration expense, $Q_{\mathrm{Cost}}=\underset{i\&Element;N_C}{\mathrm{\&Sigma;}}\left((R_{\mathrm{Qi}}+K_{\mathrm{QC}})Q_{\mathrm{Ci}}\right)+\underset{j\&Element;N_{\mathrm{SVG}}}{\mathrm{\&Sigma;}}(R_{\mathrm{SVGj}}+K_{\mathrm{SVGC}}{Q}_{\mathrm{SVGj}}),$ P _costfor annual operating cost, P _cost=η P _lossc τ, N _cthe node set of installing shunt capacitor, N _sVGthe node set of installing static reacance generator, R _qithe unit capacity installing expense of node i place shunt capacitor, K _qCthe unit capacity expense of shunt capacitor, Q _ciit is node _ithe capacity of place's installing shunt capacitor, R _sVGjthe installing expense of node j place static reacance generator, K _sVGCthe unit capacity expense of static reacance generator, Q _sVGjbe the capacity of node j place installing static reacance generator, μ is rate of return on investment, and η is power sales average electricity price, and τ is the peak load duration, and c is the regulatory factor of peak load duration, P _lossthe desired value of network loss in chance constrained programming.

Further, the probability characteristics constraints described in described reactive compensation configuration method step (3) is:

$\left\{\begin{array}{c}{P}_i-V_i\underset{j\&Element;i}{\mathrm{\&Sigma;}}{V}_j(G_{\mathrm{ij}}\cos {\mathrm{\&theta;}}_{\mathrm{ij}}+B_{\mathrm{ij}}\sin {\mathrm{\&theta;}}_{\mathrm{ij}})=0\\ Q_i-V_i\underset{j\&Element;i}{\mathrm{\&Sigma;}}{V}_j(G_{\mathrm{ij}}\sin {\mathrm{\&theta;}}_{\mathrm{ij}}-B_{\mathrm{ij}}\cos {\mathrm{\&theta;}}_{\mathrm{ij}})=0\\ P_{{\mathrm{Gk}}_{\min }}\&le;P_{\mathrm{Gk}}\&le;P_{{\mathrm{Gk}}_{\max }}\\ Q_{{\mathrm{Gk}}_{\min }}\&le;Q_{\mathrm{Gk}}\&le;Q_{{\mathrm{Gk}}_{\max }}\\ Q_{{\mathrm{ci}}_{\min }}\&le;Q_{\mathrm{ci}}\&le;Q_{{\mathrm{ci}}_{\max }}\\ T_{j_{\min }}\&le;T_j\&le;T_{j_{\max }}\\ P\{P_{k_{\min }}\&le;P_k\&le;P_{k_{\max }},k=\mathrm{1,2},...,b\}\&GreaterEqual;{\mathrm{\&alpha;}}_1\\ P\{V_{i_{\min }}\&le;V_i\&le;V_{i_{\max }},i=\mathrm{1,2},...,n\}\&GreaterEqual;{\mathrm{\&alpha;}}_2\end{array}\right.,$ Wherein:

V _jfor the voltage magnitude at node j place, θ _ijfor the voltage phase difference of node i and node j, P _gkfor generator active power of output, for the minimum value of generator active power of output, for generator active power of output maximum, Q _gkfor generator output reactive power, for the minimum value of generator output reactive power, for generator output reactive power maximum, for the minimum value of node i place installing shunt capacitor active volume, for the maximum of node i place installing shunt capacitor active volume, T _jfor adjustable transformer divides tap position, for adjustable transformer divides the minimum value of tap position, for adjustable transformer divides the maximum of tap position, P _kfor line transmission power, for circuit allows the minimum value of through-put power, for circuit allows the maximum of through-put power, b is system line sum, for the minimum value of the voltage magnitude that node i place allows, for the maximum of the voltage magnitude that node i place allows, α ₁, α ₂for the confidence bound level for probability nature.

Further, the step (4) of described reactive compensation configuration method, with the primary of idle work optimization chance constraint Mathematical Modeling for framework formation population, starts iteration and carries out Load flow calculation:

A, obtains primary population size parameter, arranges maximum iteration time, produces all particles and himself initial position and initial velocity at random,

B, brings particle each in population into power flow algorithm and obtains each circuit active power and each node voltage,

C, when circuit active power and node voltage do not meet probability characteristics constraints and/or fluctuation characteristic constraints, increases penalty term and calculates fitness in target function, otherwise, directly calculate fitness by target function;

D, more the flying speed of new particle and position, return step b until reach maximum iteration time Output rusults.

The present invention adopts technique scheme; there is following beneficial effect: the problem solving the grid-connected idle work optimization of large-scale photovoltaic power station cluster; the optimization method based on chance constraint constructed by the present invention has taken into full account photovoltaic power probability characteristics and wave characteristic; the reactive power compensation configuration result calculating gained significantly improves voltage levvl and the quality of power supply in operation of power networks, and can be applied in the idle project study of other electricity generation systems with random character and fluctuation characteristic.

Accompanying drawing explanation

Fig. 1 is the grid-connected structural representation of large-scale photovoltaic power station cluster;

Fig. 2 is the reactive-load compensation equipment installation position schematic diagram that the present invention proposes;

Fig. 3 is the photovoltaic year power fluctuation accumulated probability curve that the present invention proposes;

Fig. 4 is reactive power compensation configuration optimization calculation flow chart of the present invention;

Fig. 5 is the grid-connected generating and transmitting system legend of typical photovoltaic cluster;

Fig. 6 is embodiment beneficial outcomes.

Embodiment

Be described in detail below in conjunction with the technical scheme of accompanying drawing to invention.

Fig. 1 is shown in by large-scale photovoltaic grid connected structure schematic diagram.In order to realize the optimal compensation of such network, shunt capacitor and static reacance generator (static var generator is called for short SVG) is adopted to carry out collaborative configuration compensation.What wherein shunt capacitor was installed in generating and transmitting system collects substation low-voltage side bus, and SVG is installed in photovoltaic plant outlet bus, as shown in Figure 2.Set up based on the Optimal reactive power method of chance constraint, in the method, realized the configuration of shunt capacitor by the chance constraint of traditional grid structure, by photovoltaic year, the chance constraint of power fluctuation accumulated probability curve realizes the configuration of photovoltaic plant SVG.Finally adopt random particles group algorithm based on Latin hypercube Probabilistic Load Flow as solving mode.

Of the present invention specifically based on chance constraint idle planing method as shown in Figure 4, specifically have:

(1) form node admittance matrix in systems in which and set each node initial voltage;

(2) generator reactive power Q is chosen _gk, reactive-load compensation equipment capacity Q _ci, adjustable transformer divides tap T _jfor control variables, choose the line transmission active-power P that node injects _k, node voltage V _ifor state variable;

(3) set up idle work optimization chance constraint Mathematical Modeling, specifically comprise target function, probability characteristics constraints and fluctuation characteristic constraints,

1, target function: with generating and transmitting system equivalence annual operating cost for target, comprises two parts reactive power compensation configuration expense Q _costwith annual operating cost P _cost, computational methods are:

$\begin{array}{c}{Q}_{\mathrm{Cost}}=\underset{i\&Element;N_C}{\mathrm{\&Sigma;}}\left((R_{\mathrm{Qi}}+K_{\mathrm{QC}})Q_{\mathrm{Ci}}\right)\\ +\underset{j\&Element;N_{\mathrm{SVG}}}{\mathrm{\&Sigma;}}(R_{\mathrm{SVGj}}+K_{\mathrm{SVGC}}{Q}_{\mathrm{SVGj}})\end{array}---\left(1\right)$

P _Cost＝ηP _losscτ (2)

minf＝μQ _Cost+P _Cost(3)

In formula, N _crepresent the node set of installing shunt capacitor, R _qithat the unit capacity (every MVar) of shunt capacitor installs expense, K _qCshunt capacitor unit capacity (every MVar) expense, Q _ciit is the capacity of node i place installing shunt capacitor; N _sVGrepresent the node set of installing SVG, R _sVGjthe installing expense of SVG, K _sVGCthe unit capacity expense of SVG, Q _sVGjit is the capacity of node j place installing SVG.η is power sales average electricity price, P _lossthe desired value of network loss in chance constrained programming, τ represents the peak load duration, the regulatory factor of c peak load duration, and c τ is the continuous loading time of equivalence during chance constraint is optimized.μ is rate of return on investment.

2, probability characteristics constraints: consider from probability characteristics angle merely, build complete constraints, comprise equality constraint, certainty inequality constraints and probability inequality constraints:

$\begin{array}{c}{P}_i-V_i\underset{j\&Element;i}{\mathrm{\&Sigma;}}{V}_j(G_{\mathrm{ij}}\cos {\mathrm{\&theta;}}_{\mathrm{ij}}+B_{\mathrm{ij}}\sin {\mathrm{\&theta;}}_{\mathrm{ij}})=0\\ Q_i-V_i\underset{j\&Element;i}{\mathrm{\&Sigma;}}{V}_j(G_{\mathrm{ij}}\sin {\mathrm{\&theta;}}_{\mathrm{ij}}-B_{\mathrm{ij}}\cos {\mathrm{\&theta;}}_{\mathrm{ij}})=0\end{array}---\left(4\right)$

$\begin{array}{c}{P}_{{\mathrm{Gk}}_{\min }}\&le;P_{\mathrm{Gk}}\&le;P_{{\mathrm{Gk}}_{\max }}\\ Q_{{\mathrm{Gk}}_{\min }}\&le;Q_{\mathrm{Gk}}\&le;Q_{{\mathrm{Gk}}_{\max }}\\ Q_{{\mathrm{ci}}_{\min }}\&le;Q_{\mathrm{ci}}\&le;Q_{{\mathrm{ci}}_{\max }}\\ T_{j_{\min }}\&le;T_j\&le;T_{j_{\max }}\end{array}---\left(5\right)$

$\begin{array}{c}P\{P_{k_{\min }}\&le;P_k\&le;P_{k_{\max }},k=\mathrm{1,2},...,b\}\&GreaterEqual;{\mathrm{\&alpha;}}_1\\ P\{V_{i_{\min }}\&le;V_i\&le;V_{i_{\max }},i=\mathrm{1,2},...,n\}\&GreaterEqual;{\mathrm{\&alpha;}}_2\end{array}---\left(6\right)$

V _jfor the voltage magnitude at node j place, θ _ijfor the voltage phase difference of node i and node j, P _gkfor generator active power of output, for generator can the minimum value of active power of output, for generator can active power of output maximum, Q _gkfor generator output reactive power, for generator can the minimum value of output reactive power, for generator can output reactive power maximum, for the minimum value of node i place installing shunt capacitor active volume, for the maximum of node i place installing shunt capacitor active volume, T _jfor adjustable transformer divides tap position, for adjustable transformer divides the minimum value of tap position, for adjustable transformer divides the maximum of tap position, P _kfor line transmission power, for circuit allows the minimum value of through-put power, for circuit allows the maximum of through-put power, b is system line sum, for the voltage magnitude minimum value that node i place allows, for the voltage magnitude maximum that node i place allows, α ₁, α ₂for the confidence bound level for probability nature.

3, fluctuation characteristic constraints: in order to take into account the impact that photovoltaic power fluctuation produces system, first photovoltaic year power fluctuation accumulated probability curve is proposed, this curve definitions is as follows: using photovoltaic power fluctuation numerical value as abscissa, to be less than or equal to the cumulative probability of given fluctuation numerical value as ordinate, the curve drawn out is a year power fluctuation accumulated probability curve, as shown in Figure 3.Based on this, can obtain according to Fig. 3 the reactive configuration method considering power fluctuation feature:

$\begin{array}{c}{P}_i+\mathrm{\&Delta;}{P}_i-{V_i}^{\&prime;}\underset{j\&Element;i}{\mathrm{\&Sigma;}}{V_j}^{\&prime;}(G_{\mathrm{ij}}\cos {{\mathrm{\&theta;}}_{\mathrm{ij}}}^{\&prime;}+B_{\mathrm{ij}}\sin {{\mathrm{\&theta;}}_{\mathrm{ij}}}^{\&prime;})=0\\ Q_i+\mathrm{\&Delta;}{Q}_{{\mathrm{SVG}}_i}-{V_i}^{\&prime;}\underset{j\&Element;i}{\mathrm{\&Sigma;}}{V_j}^{\&prime;}(G_{\mathrm{ij}}\sin {{\mathrm{\&theta;}}_{\mathrm{ij}}}^{\&prime;}-B_{\mathrm{ij}}\cos {{\mathrm{\&theta;}}_{\mathrm{ij}}}^{\&prime;})=0\end{array}---\left(7\right)$

P{ΔP _i ^-≤ΔP _i≤ΔP _i ⁺,i＝1,2,...,n}＞β

(8)

abs(V _i-V _i′)＜Req,i＝1,2,...,n

P _ifor the active power that node i place is injected, Δ P _ifor the variable quantity of active power is injected at node i place, V _ifor the voltage magnitude at node i place, V _i' be the voltage magnitude at node i place during power fluctuation, V _j' be the voltage magnitude at node j place during power fluctuation, G _ijfor network conductance, B _ijfor network susceptance, θ _ij' be the voltage phase difference of node i during power fluctuation and node j, Q _ifor the reactive power that node i place is injected, for during power fluctuation, node i place reacance generator sends the variable quantity of reactive power, Δ P _i ^-for the minimum value in confidence bound level lower node i place active power fluctuation, Δ P _i ⁺for the maximum in confidence bound level lower node i place active power fluctuation, n is interstitial content, and β is the confidence bound level for wave characteristic, and abs () is signed magnitude arithmetic(al), and Req is the scope of admissible voltage fluctuation.

(4) adopt the configuration based on the random particles group algorithm determination control variables of Latin hypercube Probabilistic Load Flow, with the primary of idle work optimization chance constraint Mathematical Modeling for framework formation population, start iteration and carry out Load flow calculation:

A, obtains primary population size parameter, arranges maximum iteration time, produces all particles and himself initial position (corresponding to control variables) and initial velocity at random,

B, brings particle each in population into power flow algorithm and obtains circuit active power and each node voltage,

C, when circuit active power and node voltage do not meet probability characteristics constraints (formula 4 to formula 6) and/or fluctuation characteristic constraints (formula 7 and formula 8), in target function, increase penalty term in conjunction with formula 1 to 3 and calculate fitness, otherwise, directly calculate fitness by target function;

D, more the flying speed of new particle and position, return step b until reach maximum iteration time Output rusults.

Formula (1) constructs complete reactive compensation configuration method to (8); both the probability characteristics of photovoltaic power had been considered in the method; considering again the fluctuation characteristic of photovoltaic power, is the brand-new method being applicable to the grid-connected idle planning of large-scale photovoltaic.

According to the large-scale photovoltaic synchronizing mode of Fig. 1, set up the grid-connected figure of typical photovoltaic cluster shown in Fig. 5.In this photovoltaic cluster generating system, node 1 is and site, and the installed capacity of each photovoltaic plant is in table 1.

Node	6	8	12	14	18	20	22	24	26	28
											Capacity/MW	10	30	20	20	10	10	20	20	30	40

Table 1

After adopting the method for the invention configuration, reactive power compensation result is as table 2:

Node	2	4	5	7	10	11	13
								Type	Capacitor	Capacitor	SVG	SVG	Capacitor	SVG	SVG
Capacity/Mvar	3.00	1.00	0.15	0.44	0.50	0.30	0.30
								Node	16	17	19	21	23	25	27
Type	Capacitor	SVG	SVG	SVG	SVG	SVG	SVG
								Capacity/Mvar	2.00	0.15	0.16	0.28	0.28	0.44	0.59

Table 2

Compared with existing basic methods, in order to superiority of the present invention to be described.Basic methods only considers the computation modeling of formula (1)-(6), namely only considers the probability of photovoltaic power and does not consider fluctuation.For the 30MW photovoltaic plant at node 8 place, in basic methods, the configuration capacity of SVG is 0.30MVar, and after adopting the present invention, configuration reaches 0.44MVar, illustrates that the present invention and basic methods exist larger result difference.

After adopting the present invention, integrated planning result is shown in Fig. 6 compared to the beneficial effect of basic methods.Under two kinds of reactive power compensation allocation optimum modes, obvious the inventive method makes the voltage of node 16 fuctuation within a narrow range in less scope, and line voltage desired value is improved.Thus the present invention effectively raise reactive power compensation configuration after the electrical network quality of power supply and voltage levvl.

Claims (5)

1. reactive compensation configuration method, is characterized in that, adopts shunt capacitor and static reacance generator to carry out collaborative configuration compensation, specifically comprises the steps:

(1) in electric power system, form node admittance matrix and set each node initial voltage;

(2) control variables and state variable is chosen;

(3) set up idle work optimization chance constraint Mathematical Modeling, specifically comprise target function, probability characteristics constraints and fluctuation characteristic constraints, wherein, fluctuation constraints:

$\left\{\begin{array}{c}{P}_i+{\mathrm{\&Delta;P}}_i-V_i^{\&prime;}\underset{j\&Element;i}{\mathrm{\&Sigma;}}{V}_j^{\&prime;}(G_{\mathrm{ij}}\cos {\mathrm{\&theta;}}_{\mathrm{ij}}^{\&prime;}+B_{\mathrm{ij}}\sin {\mathrm{\&theta;}}_{\mathrm{ij}}^{\&prime;})=0\\ Q_i+{\mathrm{\&Delta;Q}}_{{\mathrm{SVG}}_i}-V_i^{\&prime;}\underset{j\&Element;i}{\mathrm{\&Sigma;}}{V}_j^{\&prime;}(G_{\mathrm{ij}}\sin {\mathrm{\&theta;}}_{\mathrm{ij}}^{\&prime;}-B_{\mathrm{ij}}\cos {\mathrm{\&theta;}}_{\mathrm{ij}}^{\&prime;})=0\\ P\{{\mathrm{\&Delta;P}}_i^{-}\&le;{\mathrm{\&Delta;P}}_i\&le;{\mathrm{\&Delta;P}}_i^{+},i=\mathrm{1,2},...,n\}>\mathrm{\&beta;}\\ \mathrm{abs}(V_i-V_i^{\&prime;})<\mathrm{Req},i=\mathrm{1,2},...,n\end{array}\right.,$ Wherein,

P _ifor the active power that node i place is injected, Δ P _ifor the variable quantity of active power is injected at node i place, V _ifor the voltage magnitude at node i place, V _i' be the voltage magnitude at node i place during power fluctuation, V _j' be the voltage magnitude at node j place during power fluctuation, G _ijfor network conductance, B _ijfor network susceptance, θ _ij' be the voltage phase difference of node i during power fluctuation and node j, Q _ifor the reactive power that node i place is injected, for during power fluctuation, node i place reacance generator sends the variable quantity of reactive power, β is the confidence bound level for wave characteristic, for the minimum value in confidence bound level lower node i place active power fluctuation, for the maximum in confidence bound level lower node i place active power fluctuation, n is interstitial content, and abs () is signed magnitude arithmetic(al), and Req is the scope of admissible voltage fluctuation.

(4) configuration based on the random particles group algorithm determination control variables of Latin hypercube Probabilistic Load Flow is adopted.

2. reactive compensation configuration method according to claim 1, it is characterized in that, choose generator reactive power in step (2), shunt capacitor capacity, adjustable transformer divide tap to be control variables, active power, the node voltage of access line transmission are state variable.

3. reactive compensation configuration method according to claim 2, is characterized in that, sets up target function f, minf=μ Q for target in step (3) so that system equivalence annual operating cost is minimum _cost+ P _cost, Q _costfor reactive power compensation configuration expense, $Q_{\mathrm{Cost}}=\underset{i\&Element;N_C}{\mathrm{\&Sigma;}}\left((R_{\mathrm{Qi}}+K_{\mathrm{QC}})Q_{\mathrm{Ci}}\right)+\underset{j\&Element;N_{\mathrm{SVG}}}{\mathrm{\&Sigma;}}(R_{{\mathrm{SVG}}_j}+K_{\mathrm{SVGC}}{Q}_{\mathrm{SVGj}}),$ P _costfor annual operating cost, P _cost=η P _lossc τ, N _cthe node set of installing shunt capacitor, N _sVGthe node set of installing static reacance generator, R _qithe unit capacity installing expense of node i place shunt capacitor, K _qCthe unit capacity expense of shunt capacitor, Q _ciit is node _ithe capacity of place's installing shunt capacitor, R _sVGjthe installing expense of node j place static reacance generator, K _sVGCthe unit capacity expense of static reacance generator, Q _sVGjbe the capacity of node j place installing static reacance generator, μ is rate of return on investment, and η is power sales average electricity price, and τ is the peak load duration, and c is the regulatory factor of peak load duration, P _lossthe desired value of network loss in chance constrained programming.

4. reactive compensation configuration method according to claim 3, is characterized in that, the probability characteristics constraints described in step (3) is:

$\left\{\begin{array}{c}{P}_i-V_i\underset{j\&Element;i}{\mathrm{\&Sigma;}}{V}_j(G_{\mathrm{ij}}\cos {\mathrm{\&theta;}}_{\mathrm{ij}}+B_{\mathrm{ij}}\sin {\mathrm{\&theta;}}_{\mathrm{ij}})=0\\ Q_i-V_i\underset{j\&Element;i}{\mathrm{\&Sigma;}}{V}_j(G_{\mathrm{ij}}\sin {\mathrm{\&theta;}}_{\mathrm{ij}}-B_{\mathrm{ij}}\cos {\mathrm{\&theta;}}_{\mathrm{ij}})=0\\ P_{{\mathrm{Gk}}_{\min }}\&le;P_{\mathrm{Gk}}\&le;P_{{\mathrm{Gk}}_{\max }}\\ Q_{{\mathrm{Gk}}_{\min }}\&le;Q_{\mathrm{Gk}}\&le;Q_{{\mathrm{Gk}}_{\max }}\\ Q_{{\mathrm{ci}}_{\min }}\&le;Q_{\mathrm{ci}}\&le;Q_{{\mathrm{ci}}_{\max }}\\ T_{j_{\min }}\&le;T_j\&le;T_{j_{\max }}\\ P\{P_{k_{\min }}\&le;P_k\&le;P_{k_{\max }},k=\mathrm{1,2},...,b\}\&GreaterEqual;{\mathrm{\&alpha;}}_1\\ P\{V_{i_{\min }}\&le;V_i\&le;V_{i_{\max }},i=\mathrm{1,2},...,n\}\&GreaterEqual;{\mathrm{\&alpha;}}_2\end{array},\right.$ Wherein:

V _jfor the voltage magnitude at node j place, θ _ijfor the voltage phase difference of node i and node j, P _gkfor generator active power of output, for the minimum value of generator active power of output, for generator active power of output maximum, Q _gkfor generator output reactive power, for the minimum value of generator output reactive power, for generator output reactive power maximum, for the minimum value of node i place installing shunt capacitor active volume, for the maximum of node i place installing shunt capacitor active volume, T _jfor adjustable transformer divides tap position, for adjustable transformer divides the minimum value of tap position, for adjustable transformer divides the maximum of tap position, P _kfor line transmission power, for circuit allows the minimum value of through-put power, for circuit allows the maximum of through-put power, b is system line sum, for the minimum value of the voltage magnitude that node i place allows, for the maximum of the voltage magnitude that node i place allows, α ₁, α ₂for the confidence bound level for probability nature.

5. according to the reactive compensation configuration method in Claims 1-4 described in any one claim, it is characterized in that, step (4), with the primary of idle work optimization chance constraint Mathematical Modeling for framework formation population, starts iteration and carries out Load flow calculation:

A, obtains primary population size parameter, arranges maximum iteration time, produces all particles and himself initial position and initial velocity at random,

B, brings particle each in population into power flow algorithm and obtains each circuit active power and each node voltage,

C, when circuit active power and node voltage do not meet probability characteristics constraints and/or fluctuation characteristic constraints, increases penalty term and calculates fitness in target function, otherwise, directly calculate fitness by target function;

D, more the flying speed of new particle and position, return step b until reach maximum iteration time Output rusults.