CN103701134A - Grid-connected wind power plant point voltage control method based on MCR (Magnetic Control Reactor) and capacitance mixed compensation - Google Patents

Abstract

The invention relates to a grid-connected wind power plant point voltage control method based on MCR (Magnetic Control Reactor) and capacitance mixed compensation. The method comprises the steps of collecting real-time data of a grid-connected wind power plant region, and receiving a voltage instruction issued by a superior AVC (Automatic Voltage Control); judging whether the grid-connected wind power plant point voltage meets a requirement, if not, taking the smallest grid-connected point voltage deviation and the smallest input of reactive-load compensation equipment as targets, calling a multi-target evolutionary particle swarm optimization to solve the current input reactive compensation, obtaining the final switching amount of each equipment according to a capacitance and MCR mixed compensation strategy, and issuing the same to finally realize the effect that the grid-connected wind power plant point voltage quality is kept at a higher level. The needed switching amount of each equipment can be accurately solved by the multi-objective optimization model and a static and dynamic reactive-load compensation equipment mixed compensation strategy, and the aims of rapidly responding to the instruction issued by the superior, stabilizing the grid-connected wind power plant point voltage and improving electric energy quality are achieved.

Description

Wind farm grid-connected point voltage control method based on MCR and electric capacity mixed compensation

Technical field

The present invention relates to a kind of wind farm grid-connected point voltage control method, especially relate to a kind of wind farm grid-connected point voltage control method based on MCR and electric capacity mixed compensation.

Background technology

Fast development along with power industry. wind energy turbine set scale increases year by year, and wind energy turbine set is also increasing on the impact of electrical network.Because the factors such as fluctuations in wind speed, electric network fault cause wind farm grid-connected point voltage, fall, can bring the transient state problems such as rise of rotational speed, overvoltage, overcurrent simultaneously, have a strong impact on the safe operation of wind-powered electricity generation unit.In this case, wind-powered electricity generation unit is taked the self-protection of passive type. off-the-line from electrical network, thus ensured the safety of wind-powered electricity generation unit.Yet the proportion of wind-powered electricity generation in electrical network is larger now, blower fan is still taked self-protection and is cut out electrical network when grid-connected point voltage falls; can have a strong impact on the stable operation of other parts in net; even cause the aggravation of fault, cause other unit trips, finally cause electrical network paralysis.Therefore must take effective wind farm grid-connected point voltage control measure, guarantee wind farm grid-connected stability.

At present the research of micro-electrical network is mainly concentrated on and utilizes the blower fan type having based on P-Q decoupling zero control to regulate as controlled continuously reactive source participation reactive voltage, and this will make the possibility of wind energy turbine set through fault reduce, and the infringement of electrical network is increased on the contrary.Therefore,, for the deficiency of research, problem research that wind farm grid-connected point voltage is controlled seems very necessary.

Summary of the invention

Above-mentioned technical problem of the present invention is mainly solved by following technical proposals:

A wind farm grid-connected point voltage control method based on MCR and electric capacity mixed compensation, is characterized in that: comprise the following steps:

Step 1, gathers wind farm grid-connected region real time data, and described real time data comprises that wind energy turbine set is exerted oneself in real time, higher level's electrical network equivalent voltage and Real-time Load, and receives higher level AVC and issue voltage instruction;

Step 2, judges whether grid-connected point voltage meets the demands, if voltage deviation higher level command voltage permissible error is interval, voltage meets the demands, and proceeds to step 1; If do not meet the demands, proceed to step 3;

Step 3, sets up and take grid-connected point voltage and depart from the website idle work optimization model that minimum, reactive-load compensation equipment input amount minimum are target, establishes system restriction condition, with state variable constrain, sets up penalty function;

Step 4, website idle work optimization model and constraints based on step 3 foundation, adopt multi-target evolution Chaos particle swarm optimization algorithm to carry out iterative computation, obtains non-domination disaggregation, and calculate optimum switching combination;

Step 5, the optimum switching obtaining based on step 4 combination, and draw finally each equipment switching amount according to capacitor group and MCR mixing control strategy, and issuing, finally realizes wind farm grid-connected point voltage and meets fast higher level AVC and issue index.

In the above-mentioned wind farm grid-connected point voltage control method based on MCR and electric capacity mixed compensation, described step 3 website Optimized model is based on following formula:

$\mathrm{<figure-callout\; id="1"\; label="min\; F"\; filenames="1401091659203.png"\; state="\{\{state\}\}">min</figure-callout>}{F}_{}{}_1=Q_C/Q_{C\max }+Q_L/Q_{L\max }+\mathrm{\&lambda;}\underset{i=1}{\overset{N}{\mathrm{\&Sigma;}}}{\left(\frac{{\mathrm{\&Delta;V}}_i}{\mathrm{\&Delta;V}}\right)}^2$

$\min {\mathrm{<figure-callout\; id="2"\; label="F"\; filenames="1401091659203.png"\; state="\{\{state\}\}">F</figure-callout>}}_2={\mathrm{\&Delta;V}}_{\mathrm{pcc}}/{\mathrm{\&Delta;<figure-callout\; id="0"\; label="V\; pcc"\; filenames="1401091659203.png"\; state="\{\{state\}\}">V</figure-callout>}}_{\mathrm{pcc}}+\mathrm{\&lambda;}\underset{i=1}{\overset{N}{\mathrm{\&Sigma;}}}{\left(\frac{{\mathrm{\&Delta;V}}_i}{\mathrm{\&Delta;V}}\right)}^2$

Wherein,

$\left\{\begin{array}{c}{\mathrm{\&Delta;V}}_i=\left\{\begin{array}{cc}{V}_i-V_{i\max }& V_i>V_{i\max }\\ 0& V_{i\min }\&le;V_i\&le;V_{i\max }\\ V_{i\min }-V_i& V_i<V_{i\min }\end{array}\right.\\ \mathrm{\&Delta;V}=V_{i\max }-V_{i\min }\\ \mathrm{\&Delta;}{V}_{\mathrm{pcc}}={(V_{\mathrm{pcc}}-V_{\mathrm{pccref}})}^2\\ {\mathrm{\&Delta;V}}_{\mathrm{<figure-callout\; id="0"\; label="pcc"\; filenames="1401091659203.png"\; state="\{\{state\}\}">pcc</figure-callout>}0}={(V_{\mathrm{pcc}0}-V_{\mathrm{pccref}})}^2\end{array}\right.$

In formula, F ₁and F ₂for optimization aim, F ₁for reactive-load compensation equipment input amount index, F ₂for grid-connected point voltage divergence indicator, F ₁and F ₂in

be the out-of-limit penalty of node voltage, λ is the node voltage penalty coefficient that crosses the border; V _imin, V _i, V _imaxbe respectively node voltage lower limit, node voltage, the node voltage upper limit; N is node sum; V _pccfor compensating rear PCC voltage, V _pccreffor higher level AVC issues PCC voltage, V _pcc0pCC voltage when not compensating.

In the above-mentioned wind farm grid-connected point voltage control method based on MCR and electric capacity mixed compensation, in described step 3, the constraints in wind farm grid-connected region is: and the adjustable variable of web site idle work optimization is divided into control variables and state variable, wherein capacitor and MCR input capacity are control variables, system node voltage V is state variable, based on following formula:

Constraints one, the inequality constraints of control variables:

$\left\{\begin{array}{c}{Q}_{C\min }\&le;Q_C\&le;Q_{C\max }\\ Q_{L\min }\&le;Q_L\&le;Q_{L\max }\end{array}\right.$

Constraints two, the inequality constraints of state variable is:

V _imin≤V _i≤V _imax i=1,2,…,N

Q _cmax, Q _cminand Q _cbe respectively and the idle upper and lower limit of web site capacitive compensation and the actual capacitive reactive power compensating; Q _lmax, Q _lminand Q _lbe respectively and the perception of web site perception compensating reactive power bound and actual compensation idle; N is grid-connected Area Node sum, V _imax, V _iminand V _ibe respectively i node qualified voltage bound and actual node voltage.

In the above-mentioned wind farm grid-connected point voltage control method based on MCR and electric capacity mixed compensation, described step 4 multi-target evolution Chaos particle swarm optimization algorithm flow process is:

Step 4.1, sets iterations, population number, and based on random position and the speed assignment of giving each particle of constraints one, position and the speed assignment of giving at random particle, complete the initialization of population;

Step 4.2, calculate the fitness value of each particle, do not meet constraints two, with penalty function form, count in target function, the fitness value of initialization particle, individual optimal value, obtain initial Noninferior Solution Set according to Pareto theory, and in Noninferior Solution Set, select global optimum's particle;

Step 4.3, iterations adds one, and more the position of new particle and speed, calculate new fitness;

Step 4.4, merges all particles and last iteration gained Noninferior Solution Set, based on pareto theory, obtains Noninferior Solution Set of new generation, and part noninferior solution is carried out to chaotic mutation;

Step 4.5, when reaching maximum iteration time, iteration stopping, exports final Noninferior Solution Set; Otherwise forward step 4.3 to until finish after reaching maximum iteration time.

In the above-mentioned wind farm grid-connected point voltage control method based on MCR and electric capacity mixed compensation, described step 5 capacitor group and MCR mixing control strategy are based on following principle: electric capacity is discrete type capacitive reactive power compensation equipment, can only classification switching, MCR is continuous type reactive-load compensation equipment, therefore the two is coordinated to control and can reach idle continuous fine adjustment, and concrete control method is based on following formula:

Q=kQ _c0+Q _MCR,

Wherein, Q is for mixing the idle total amount of switching, Q _c0for single group capacitor capacity, k is for dropping into capacitance group number, Q _sVGfor MCR drops into idle amount, its adjustable range is [kQ _c0, 0];

Define while controlling under the condition that last voltage meets the demands and drop into k group electric capacity, according to the judged result of this gained capacitive reactive power Q, select to carry out:

Select to carry out one: this gained capacitive reactive power Q meets (k-1) Q _c0<Q<kQ _c0, this controls the final k of input group electric capacity, and MCR input capacity is Q-(k-1) Q _c0perception idle;

Select to carry out two: this gained capacitive reactive power Q meets kQ _c0<Q< (k+1) Q _c0, and current PCC voltage deviation is while controlling dead band, and this controls the final k+1 of input group electric capacity, and MCR input capacity is (k+1) Q _c0the perception of-Q is idle,

Select to carry out three: this gained capacitive reactive power Q meets kQ _c0<Q< (k+1) Q _c0, and current PCC voltage do not depart from and controls dead band, and this controls the final k of input group electric capacity, and it is 0 that MCR drops into perceptual idle amount.

Therefore, tool of the present invention has the following advantages: 1. introduced grid-connected point voltage and departed from reactive apparatus input amount as optimization aim, both guaranteed the stable of grid-connected point voltage, made again equipment switching amount reach minimum; 2. introduce chaotic mutation, improved the ability of searching optimum of particle cluster algorithm; 3. introduce penalty function, eliminate nonconforming solution; 4. the capacitor group adopting can reduce capacitor switching number of times effectively with MCR mixing control strategy, and the grid-connected point voltage of quick adjustment makes it to meet under higher level AVC and sends instructions, and reaches the object of stablizing grid-connected point voltage.

Accompanying drawing explanation

Fig. 1 is process chart of the present invention.

Fig. 2 is the electric hookup in the embodiment of the present invention.

Fig. 3 be not in the same time blower fan exert oneself and duty ratio illustration.

Fig. 4 is grid-connected point voltage correlation curve figure before and after single-throw Capacity control.

Fig. 5 is that electric capacity mixes the grid-connected point voltage correlation curve figure in control front and back with MCR.

Embodiment

Below by embodiment, and by reference to the accompanying drawings, technical scheme of the present invention is described in further detail.

Embodiment:

One, paper concrete steps of the present invention once, mainly comprise:

Step 1, gathers wind farm grid-connected region real time data, and described real time data comprises that wind energy turbine set is exerted oneself in real time, higher level's electrical network equivalent voltage and Real-time Load, and receives higher level AVC and issue voltage instruction;

Step 2, judges whether grid-connected point voltage meets the demands, if voltage deviation higher level command voltage permissible error is interval, voltage meets the demands, and proceeds to step 1; If do not meet the demands, proceed to step 3;

Step 3, sets up and take grid-connected point voltage and depart from the website idle work optimization model that minimum, reactive-load compensation equipment input amount minimum are target, establishes system restriction condition, with state variable constrain, sets up penalty function;

1. wherein, website Optimized model is based on following formula:

$\mathrm{<figure-callout\; id="1"\; label="min\; F"\; filenames="1401091659203.png"\; state="\{\{state\}\}">min</figure-callout>}{F}_{}{}_1=Q_C/Q_{C\max }+Q_L/Q_{L\max }+\mathrm{\&lambda;}\underset{i=1}{\overset{N}{\mathrm{\&Sigma;}}}{\left(\frac{{\mathrm{\&Delta;V}}_i}{\mathrm{\&Delta;V}}\right)}^2$

$\min {\mathrm{<figure-callout\; id="2"\; label="F"\; filenames="1401091659203.png"\; state="\{\{state\}\}">F</figure-callout>}}_2={\mathrm{\&Delta;V}}_{\mathrm{pcc}}/{\mathrm{\&Delta;<figure-callout\; id="0"\; label="V\; pcc"\; filenames="1401091659203.png"\; state="\{\{state\}\}">V</figure-callout>}}_{\mathrm{pcc}}+\mathrm{\&lambda;}\underset{i=1}{\overset{N}{\mathrm{\&Sigma;}}}{\left(\frac{{\mathrm{\&Delta;V}}_i}{\mathrm{\&Delta;V}}\right)}^2$

Wherein,

$\left\{\begin{array}{c}{\mathrm{\&Delta;V}}_i=\left\{\begin{array}{cc}{V}_i-V_{i\max }& V_i>V_{i\max }\\ 0& V_{i\min }\&le;V_i\&le;V_{i\max }\\ V_{i\min }-V_i& V_i<V_{i\min }\end{array}\right.\\ \mathrm{\&Delta;V}=V_{i\max }-V_{i\min }\\ \mathrm{\&Delta;}{V}_{\mathrm{pcc}}={(V_{\mathrm{pcc}}-V_{\mathrm{pccref}})}^2\\ {\mathrm{\&Delta;V}}_{\mathrm{<figure-callout\; id="0"\; label="pcc"\; filenames="1401091659203.png"\; state="\{\{state\}\}">pcc</figure-callout>}0}={(V_{\mathrm{pcc}0}-V_{\mathrm{pccref}})}^2\end{array}\right.$

In formula, F ₁and F ₂for optimization aim, F ₁for reactive-load compensation equipment input amount index, F ₂for grid-connected point voltage divergence indicator, F ₁and F ₂in

be the out-of-limit penalty of node voltage, λ is the node voltage penalty coefficient that crosses the border; V _imin, V _i, V _imaxbe respectively node voltage lower limit, node voltage, the node voltage upper limit; N is node sum; V _pccfor compensating rear PCC voltage, V _pccreffor higher level AVC issues PCC voltage, V _pcc0pCC voltage when not compensating.

2. the constraints in wind farm grid-connected region is: and the adjustable variable of web site idle work optimization is divided into control variables and state variable, and wherein capacitor and MCR input capacity are control variables, and system node voltage V is state variable, based on following formula:

Constraints one, the inequality constraints of control variables:

$\left\{\begin{array}{c}{Q}_{C\min }\&le;Q_C\&le;Q_{C\max }\\ Q_{L\min }\&le;Q_L\&le;Q_{L\max }\end{array}\right.$

Constraints two, the inequality constraints of state variable is:

V _imin≤V _i≤V _imax i=1,2,…,N

Q _cmax, Q _cminand Q _cbe respectively and the idle upper and lower limit of web site capacitive compensation and the actual capacitive reactive power compensating; Q _lmax, Q _lminand Q _lbe respectively and the perception of web site perception compensating reactive power bound and actual compensation idle; N is grid-connected Area Node sum, V _imax, V _iminand V _ibe respectively i node qualified voltage bound and actual node voltage.

Step 4, website idle work optimization model and constraints based on step 3 foundation, adopt multi-target evolution Chaos particle swarm optimization algorithm to carry out iterative computation, obtains non-domination disaggregation, and calculate optimum switching combination; Multi-target evolution Chaos particle swarm optimization algorithm flow process is:

Step 4.1, sets iterations, population number, and based on random position and the speed assignment of giving each particle of constraints one, position and the speed assignment of giving at random particle, complete the initialization of population;

Step 4.2, calculate the fitness value of each particle, do not meet constraints two, with penalty function form, count in target function, the fitness value of initialization particle, individual optimal value, obtain initial Noninferior Solution Set according to Pareto theory, and in Noninferior Solution Set, select global optimum's particle;

Step 4.3, iterations adds one, and more the position of new particle and speed, calculate new fitness;

Step 4.4, merges all particles and last iteration gained Noninferior Solution Set, based on pareto theory, obtains Noninferior Solution Set of new generation, and part noninferior solution is carried out to chaotic mutation;

Step 4.5, when reaching maximum iteration time, iteration stopping, exports final Noninferior Solution Set; Otherwise forward step 4.3 to until finish after reaching maximum iteration time.

Step 5, the optimum switching obtaining based on step 4 combination, and draw finally each equipment switching amount according to capacitor group and MCR mixing control strategy, and issuing, finally realizes wind farm grid-connected point voltage and meets fast higher level AVC and issue index; This capacitor group and MCR mixing control strategy are based on following principle: electric capacity is discrete type capacitive reactive power compensation equipment, can only classification switching, MCR is continuous type reactive-load compensation equipment, so the two coordinate to control and can reach idle continuous fine adjustment, and concrete control method is based on following formula:

Q=kQ _c0+Q _MCR,

Wherein, Q is for mixing the idle total amount of switching, Q _c0for single group capacitor capacity, k is for dropping into capacitance group number, Q _sVGfor MCR drops into idle amount, its adjustable range is [kQ _c0, 0];

Define while controlling under the condition that last voltage meets the demands and drop into k group electric capacity, according to the judged result of this gained capacitive reactive power Q, select to carry out:

Select to carry out one: this gained capacitive reactive power Q meets (k-1) Q _c0<Q<kQ _c0, this controls the final k of input group electric capacity, and MCR input capacity is Q-(k-1) Q _c0perception idle;

Select to carry out two: this gained capacitive reactive power Q meets kQ _c0<Q< (k+1) Q _c0, and current PCC voltage deviation is while controlling dead band, and this controls the final k+1 of input group electric capacity, and MCR input capacity is (k+1) Q _c0the perception of-Q is idle,

Select to carry out three: this gained capacitive reactive power Q meets kQ _c0<Q< (k+1) Q _c0, and current PCC voltage do not depart from and controls dead band, and this controls the final k of input group electric capacity, and it is 0 that MCR drops into perceptual idle amount.

Two, below in conjunction with the case that provides a concrete employing said method: as shown in Figure 1,

Step 1: gather wind farm grid-connected region real time data, comprise that wind energy turbine set is exerted oneself in real time, higher level's electrical network equivalent voltage and Real-time Load; Receive higher level AVC and issue voltage instruction.

In the present embodiment, adopt certain wind farm grid-connected region example of calculation shows, as shown in Figure 2, load capacity 14MW, the specified total capacity of wind energy turbine set being comprised of double-fed wind power generator is 84MW, wind farm grid-connected power station rated capacity is 60MW.

Step 2: judge whether grid-connected point voltage meets the demands, if voltage deviation is controlled dead band, proceed to step 1, if do not meet the demands, proceed to step 3.

Step 3: set up and take grid-connected point voltage and depart from the website idle work optimization model that minimum, reactive-load compensation equipment input amount minimum are target.。

According to step 3 website Optimized model described in formula, be:

$\left\{\begin{array}{c}\mathrm{<figure-callout\; id="1"\; label="min\; F"\; filenames="1401091659203.png"\; state="\{\{state\}\}">min</figure-callout>}{F}_{}{}_1=Q_C/Q_{C\max }+Q_L/Q_{L\max }+\mathrm{\&lambda;}\underset{i=1}{\overset{N}{\mathrm{\&Sigma;}}}{\left(\frac{{\mathrm{\&Delta;V}}_i}{\mathrm{\&Delta;V}}\right)}^2\\ \mathrm{<figure-callout\; id="2"\; label="min\; F"\; filenames="1401091659203.png"\; state="\{\{state\}\}">min</figure-callout>}{F}_{}{}_2={\mathrm{\&Delta;V}}_{\mathrm{pcc}}/{\mathrm{\&Delta;<figure-callout\; id="0"\; label="V\; pcc"\; filenames="1401091659203.png"\; state="\{\{state\}\}">V</figure-callout>}}_{\mathrm{pcc}}+\mathrm{\&lambda;}\underset{i=1}{\overset{N}{\mathrm{\&Sigma;}}}{\left(\frac{{\mathrm{\&Delta;V}}_i}{\mathrm{\&Delta;V}}\right)}^2\end{array}\right.---\left(3\right)$

In formula, F ₁and F ₂for optimization aim, F ₁for reactive-load compensation equipment input amount index, F ₂for grid-connected point voltage divergence indicator, F ₁and F ₂in last be the 3rd, 4 for the out-of-limit punishment of node voltage, λ is the node voltage penalty coefficient that crosses the border; V _imin, V _i, V _imaxfor node voltage lower limit, node voltage, the node voltage upper limit; N is node sum; V _pccfor compensating rear PCC voltage, V _pccreffor higher level AVC issues PCC voltage, V _pcc0pCC voltage when not compensating.

In the present embodiment, higher level AVC issues control command V _pccrefbe 1.0.

Step 4: establish system restriction condition, set up penalty function with state variable constrain.

Constraints comprises following content:

1. according to the controlled variable bound of formula (1):

$\left\{\begin{array}{c}{Q}_{\mathrm{mcr}\min }\&le;Q_{\mathrm{mcr}}\&le;Q_{\mathrm{mcr}\max }\\ C_{k\min }\&le;C_k\&le;C_{k\max }\end{array}\right.k=\mathrm{1,2},...,N_C---\left(1\right)$

In formula, Q _mcrfor the pondage of MCR, k is can switching capacitance group number, C _kit is the reactive compensation capacity of k switched capacitors.

In the present embodiment, control variables Q _mcrbe constrained to 0≤Q _mcr≤ 9Mvar, control variables C _kbe constrained to 0≤C _k≤ 4.5Mvar.

2. according to formula (2), obtain state variable constrain:

V _imin≤V _i≤V _imax i=1,2,…,N （2）

In formula, V _ifor node _ivoltage magnitude.

In the present embodiment, state variable V _iperunit value be constrained to 0.95≤V _gi≤ 1.05

Step 5: adopt multi-target evolution Chaos particle swarm optimization algorithm to carry out iterative computation, obtain non-domination disaggregation, and calculate optimum switching combination.

Step 6: draw final each equipment switching amount according to capacitor group and MCR mixing control strategy, and issue, finally realize wind farm grid-connected point voltage and meet fast higher level AVC and issue index.

In the present embodiment, use software emulation to control wind farm grid-connected point voltage, before and after obtaining single-throw Capacity control the different times of running shown in Fig. 3, grid-connected point voltage curve as shown in Figure 4, obtains electric capacity and MCR and jointly controls the grid-connected point voltage curve of front and back micro-grid system as shown in Figure 5.

Specific embodiment described herein is only to the explanation for example of the present invention's spirit.Those skilled in the art can make various modifications or supplement or adopt similar mode to substitute described specific embodiment, but can't depart from spirit of the present invention or surmount the defined scope of appended claims.

Claims (5)

1. the wind farm grid-connected point voltage control method based on MCR and electric capacity mixed compensation, is characterized in that: comprise the following steps:

Step 1, gathers wind farm grid-connected region real time data, and described real time data comprises that wind energy turbine set is exerted oneself in real time, higher level's electrical network equivalent voltage and Real-time Load, and receives higher level AVC and issue voltage instruction;

Step 2, judges whether grid-connected point voltage meets the demands, if voltage deviation higher level command voltage permissible error is interval, voltage meets the demands, and proceeds to step 1; If do not meet the demands, proceed to step 3;

Step 3, sets up and take grid-connected point voltage and depart from the website idle work optimization model that minimum, reactive-load compensation equipment input amount minimum are target, establishes system restriction condition, with state variable constrain, sets up penalty function;

Step 4, website idle work optimization model and constraints based on step 3 foundation, adopt multi-target evolution Chaos particle swarm optimization algorithm to carry out iterative computation, obtains non-domination disaggregation, and calculate optimum switching combination;

Step 5, the optimum switching obtaining based on step 4 combination, and draw finally each equipment switching amount according to capacitor group and MCR mixing control strategy, and issuing, finally realizes wind farm grid-connected point voltage and meets fast higher level AVC and issue index.

2. the wind farm grid-connected point voltage control method based on MCR and electric capacity mixed compensation according to claim 1, is characterized in that: described step 3 website Optimized model is based on following formula:

$\min F_1=Q_C/Q_{C\max }+Q_L/Q_{L\max }+\mathrm{\&lambda;}\underset{i=1}{\overset{N}{\mathrm{\&Sigma;}}}{\left(\frac{{\mathrm{\&Delta;V}}_i}{\mathrm{\&Delta;V}}\right)}^2$

$\min F_2={\mathrm{\&Delta;V}}_{\mathrm{pcc}}/{\mathrm{\&Delta;V}}_{\mathrm{pcc}0}+\mathrm{\&lambda;}\underset{i=1}{\overset{N}{\mathrm{\&Sigma;}}}{\left(\frac{{\mathrm{\&Delta;V}}_i}{\mathrm{\&Delta;V}}\right)}^2$

Wherein,

$\left\{\begin{array}{c}{\mathrm{\&Delta;V}}_i=\left\{\begin{array}{cc}{V}_i-V_{i\max }& V_i>V_{i\max }\\ 0& V_{i\min }\&le;V_i\&le;V_{i\max }\\ V_{i\min }-V_i& V_i<V_{i\min }\end{array}\right.\\ \mathrm{\&Delta;V}=V_{i\max }-V_{i\min }\\ \mathrm{\&Delta;}{V}_{\mathrm{pcc}}={(V_{\mathrm{pcc}}-V_{\mathrm{pccref}})}^2\\ {\mathrm{\&Delta;V}}_{\mathrm{pcc}0}={(V_{\mathrm{pcc}0}-V_{\mathrm{pccref}})}^2\end{array}\right.$

In formula, F ₁and F ₂for optimization aim, F ₁for reactive-load compensation equipment input amount index, F ₂for grid-connected point voltage divergence indicator, F ₁and F ₂in

be the out-of-limit penalty of node voltage, λ is the node voltage penalty coefficient that crosses the border; V _imin, V _i, V _imaxbe respectively node voltage lower limit, node voltage, the node voltage upper limit; N is node sum; V _pccfor compensating rear PCC voltage, V _pccreffor higher level AVC issues PCC voltage, V _pcc0pCC voltage when not compensating.

3. the wind farm grid-connected point voltage control method based on MCR and electric capacity mixed compensation according to claim 1, it is characterized in that: in described step 3, the constraints in wind farm grid-connected region is: and the adjustable variable of web site idle work optimization is divided into control variables and state variable, wherein capacitor and MCR input capacity are control variables, system node voltage V is state variable, based on following formula:

Constraints one, the inequality constraints of control variables:

$\left\{\begin{array}{c}{Q}_{C\min }\&le;Q_C\&le;Q_{C\max }\\ Q_{L\min }\&le;Q_L\&le;Q_{L\max }\end{array}\right.$

Constraints two, the inequality constraints of state variable is:

V _imin≤V _i≤V _imax i=1,2,…,N

Q _cmax, Q _cminand Q _cbe respectively and the idle upper and lower limit of web site capacitive compensation and the actual capacitive reactive power compensating; Q _lmax, Q _lminand Q _lbe respectively and the perception of web site perception compensating reactive power bound and actual compensation idle; N is grid-connected Area Node sum, V _imax, V _iminand V _ibe respectively i node qualified voltage bound and actual node voltage.

4. the wind farm grid-connected point voltage control method based on MCR and electric capacity mixed compensation according to claim 1, is characterized in that: described step 4 multi-target evolution Chaos particle swarm optimization algorithm flow process is:

Step 4.1, sets iterations, population number, and based on random position and the speed assignment of giving each particle of constraints one, position and the speed assignment of giving at random particle, complete the initialization of population;

Step 4.2, calculate the fitness value of each particle, do not meet constraints two, with penalty function form, count in target function, the fitness value of initialization particle, individual optimal value, obtain initial Noninferior Solution Set according to Pareto theory, and in Noninferior Solution Set, select global optimum's particle;

Step 4.3, iterations adds one, and more the position of new particle and speed, calculate new fitness;

Step 4.4, merges all particles and last iteration gained Noninferior Solution Set, based on Pareto theory, obtains Noninferior Solution Set of new generation, and part noninferior solution is carried out to chaotic mutation;

Step 4.5, when reaching maximum iteration time, iteration stopping, exports final Noninferior Solution Set; Otherwise forward step 4.3 to until finish after reaching maximum iteration time.

5. the wind farm grid-connected point voltage control method based on MCR and electric capacity mixed compensation according to claim 1, it is characterized in that: described step 5 capacitor group and MCR mixing control strategy are based on following principle: electric capacity is discrete type capacitive reactive power compensation equipment, can only classification switching, MCR is continuous type reactive-load compensation equipment, therefore the two is coordinated to control and can reach idle continuous fine adjustment, and concrete control method is based on following formula:

Q=kQ _c0+Q _MCR,

Wherein, Q is for mixing the idle total amount of switching, Q _c0for single group capacitor capacity, k is for dropping into capacitance group number, Q _sVGfor MCR drops into idle amount, its adjustable range is [kQ _c0, 0];

Define while controlling under the condition that last voltage meets the demands and drop into k group electric capacity, according to the judged result of this gained capacitive reactive power Q, select to carry out:

Select to carry out one: this gained capacitive reactive power Q meets (k-1) Q _c0<Q<kQ _c0, this controls the final k of input group electric capacity, and MCR input capacity is Q-(k-1) Q _c0perception idle;

Select to carry out two: this gained capacitive reactive power Q meets kQ _c0<Q< (k+1) Q _c0, and current PCC voltage deviation is while controlling dead band, and this controls the final k+1 of input group electric capacity, and MCR input capacity is (k+1) Q _c0the perception of-Q is idle,

Select to carry out three: this gained capacitive reactive power Q meets kQ _c0<Q< (k+1) Q _c0, and current PCC voltage do not depart from and controls dead band, and this controls the final k of input group electric capacity, and it is 0 that MCR drops into perceptual idle amount.

Images