CN104810863A - Generator set active power real-time dispatching method considering wind power prediction error - Google Patents

Abstract

The invention discloses an electric system active power real-time dispatching method considering wind power prediction error. The electric system active power real-time dispatching method comprises the following steps of describing random distribution character of the wind power prediction error by using Laplace distribution; by taking economic efficiency optimum and wind curtailment minimum as two targets, introducing a chance constraint condition; solving active power flow probability distribution of a system on the basis of a probabilistic power flow method for Latin hypercube sampling; establishing a buffer generator set active power optimum allocation model considering the wind power prediction error, and solving a chance constraint planning model by adopting an improved genetic algorithm. According to the electric system active power real-time dispatching method disclosed by the invention, the wind power prediction error is considered during real-time dispatching, power unbalance and flow off-limit brought to the system by wind power output offset can be avoided, and the safe operation of the system is guaranteed; moreover, while the wind power prediction error is assimilated, the wind power accepting ability is increased by the system.

Description

A kind ofly consider that the unit of wind-powered electricity generation predicated error is gained merit real-time scheduling method

Technical field

The present invention relates to the operation control field of electrical network, the particularly dispatching method of wind power new energy electrical network.

Background technology

In recent years, along with the adjustment dynamics of domestic power supply architecture continues to increase, generation of electricity by new energy technology develops rapidly, and the capacity of access electrical network increases.But the randomness that wind power generation is intrinsic and fluctuation, make it when extensive connecting system, can cause to have a strong impact on to operation of power networks and even jeopardize power grid security.

Economic Dispatch problem for target, meets the optimization problem of system power balance and operation constraints with economical optimum.In recent years, according to the thinking of " multilevel coordination, step by step refinement ", control is divided into four-stage: scheduling a few days ago, rolling scheduling, Real-Time Scheduling and AGC (automatic generation control).Real-Time Scheduling take 15min as the cycle, get functional in system, creep speed faster unit be buffering unit, according to the exerting oneself and shift to an earlier date 15min and be handed down to corresponding unit of information of forecasting adjustment buffering unit of in advance 15min, stabilize meritorious amount of unbalance, to improve the economy of operation.Due to after rolling scheduling, the amount of the required unit output revised of Real-Time Scheduling is relatively little.Real-Time Scheduling is further check on rolling scheduling basis and correction, is the optimization problems to buffering unit output in essence.Because existing wind-powered electricity generation precision of prediction is lower, when large-scale wind power is grid-connected, the fluctuate power shortage that causes of wind-powered electricity generation will bring threat to the safe operation of electrical network.But the deviation of load prediction is only considered in the research of existing research Real-Time Scheduling mostly, have ignored wind-powered electricity generation predicated error.Although load prediction also exists error, compared to wind-powered electricity generation prediction, its variation tendency is thought can by accurately predicting.In fact, there is larger difference between the measured value of wind-powered electricity generation prediction and predicted value, Real-Time Scheduling result and system safety certainly will be affected.So, the impact of wind-powered electricity generation predicated error must be considered in Real-Time Scheduling.

Summary of the invention

Goal of the invention: in order to overcome the deficiencies in the prior art, the invention provides and a kind ofly consider that the unit of wind-powered electricity generation predicated error is gained merit real-time scheduling method, the technical problem that during for solving Real-Time Scheduling in existing Economic Dispatch, wind-powered electricity generation predicated error is larger.

Technical scheme: for achieving the above object, the technical solution used in the present invention is:

Consider that the unit of wind-powered electricity generation predicated error is gained merit a real-time scheduling method, comprise the following steps that order performs:

Step one, read and wait to dispatch the network data of electric power system and unit parameter, and the unit output after up-to-date rolling scheduling in electric power system;

Step 2, setting buffering unit optimal allocation model target function of gaining merit are

$\min (\underset{v\∈N_{\mathrm{rt}}}{\mathrm{\Σ}}{a}_v^{\mathrm{rt}}\·\mathrm{\Δ}{P}_v^{\mathrm{rt}}\left(t\right)+\underset{v\∈N_{\mathrm{wind}}}{\mathrm{\Σ}}{\mathrm{\κ}}_v\·\mathrm{\Δ}{P}_v^{\mathrm{loss}}\left(t\right))---\left(1\right)$

(1) in formula:

V represents that unit comprises buffering unit and Wind turbines, N _rtthe set of buffering unit, N _windthe set of Wind turbines, buffering unit generated output unit Setup Cost, cushion unit at t adjustment amount of exerting oneself (calculate in the t-1 moment and issue) just, wind turbines t just abandon air quantity, κ _vfor abandoning wind weight factor;

Step 3, obtain the wind-powered electricity generation prediction in next cycle and exert oneself and total load information of forecasting, and threshold value P is set _wind; Utilize the wind-powered electricity generation in next cycle to predict to exert oneself the wind power output deducting current period to obtain wind-powered electricity generation gross capability prediction increment, judge that wind-powered electricity generation gross capability predicts whether increment is greater than P _wind; If be more than or equal to P _wind, then next cycle power Systematic selection abandons wind; If be less than P _wind, then next cycle power Systematic selection all receives wind power output, namely

Step 4, setting with for decision variable, form chromosome to be asked;

Step 5, random generation, just for chromosome, obtain wind-powered electricity generation predicated error sample based on Latin Hypercube Sampling, obtain electric power system effective power flow probability distribution by Probabilistic Load Flow method.Said process is specially: obtain sampling matrix after sampling according to the wind-powered electricity generation predicated error stochastic variable of Latin hypercube stochastical sampling method to every Fans; After the load that each row wind-powered electricity generation predicated error vector in sampling matrix is considered as bearing is counted respective nodes, carry out DC power flow calculating, obtained the probability distribution of circuit effective power flow by mathematical statistics.Then, whether checking meets the overall units limits of buffering unit, Line Flow constraint, buffering unit output adjustment bound constraint and reserve capacity constraint totally 4 constraintss, represents respectively by following 4 formulas;

$\underset{v\∈N_{\mathrm{rt}}}{\mathrm{\Σ}}\mathrm{\Δ}{P}_v^{\mathrm{rt}}\left(t\right)+\underset{v\∈N_{\mathrm{wind}}}{\mathrm{\Σ}}(\mathrm{\Δ}{P}_v^{\mathrm{wind}}\left(t\right)-\mathrm{\Δ}{P}_v^{\mathrm{loss}}\left(t\right))-\mathrm{\Δ}{P}_{\mathrm{load}}=0---\left(2\right)$

$\begin{array}{cc}\mathrm{Pr}\{\underset{\‾}{P_{\mathrm{Tk}}}\≤P_{\mathrm{Tk}}\left({\mathrm{\ϵ}}_v^f\left(t\right)\right)\≤\underset{\‾}{\overset{\‾}{P_{\mathrm{Tk}}}}\}\≥1-{\mathrm{\θ}}_1^{\mathrm{rt}}& \left(3\right)\\ -\mathrm{\Δ}\underset{\‾}{P_v^{\mathrm{rt}}}\≤\mathrm{\Δ}{P}_v^{\mathrm{rt}}\left(t\right)\≤\mathrm{\Δ}{P}_v^{\mathrm{rt}}& \left(4\right)\end{array}$

load prediction increment deducts current period total load for next cycle total load information of forecasting;

(3) formula is chance constraint, wherein: for the wind-powered electricity generation predicated error amount of Wind turbines, P _tkfor the active power probability distribution of the line section k after t Real-Time Scheduling wind-powered electricity generation predicated error, for the active power transfer upper limit of line section k, for the active power transfer lower limit of line section k, Pr{} represents probability, for given trend constraint confidence level, and this formula all need be set up section k arbitrary in electrical network;

(4) in formula: for the lower limit of buffering unit output adjustment, and $\mathrm{\Δ}\underset{\‾}{P_v^{\mathrm{rt}}}=\min (P_v^{\mathrm{rt}}(t-1)-P_{v,\min },\mathrm{\Δ}{P}_v^{\mathrm{dn}}),$ Wherein P _{v, min}for buffering unit output lower limit, for cushioning the maximum downward amount (being determined by buffering unit climbing capacity) of unit output, for the upper limit of buffering unit output adjustment, and wherein P _{v, max}for buffering unit v exerts oneself the upper limit, for cushioning the maximum rise amount (being determined by buffering unit climbing capacity) of unit output;

(5) formula is chance constraint, wherein: for given spinning reserve constraint confidence level; Buffering unit needs to stay sufficient reserve capacity and to dissolve wind-powered electricity generation predicated error the power-balance of guarantee system.

If meet above-mentioned institute Prescribed Properties, then the first of correspondence is put into just for population for chromosome; Otherwise regenerate chromosome and repeat above-mentioned sampling and verification method on new chromosomal basis, the chromosome of the correspondence meeting constraints being put into just for population, until just reach population scale N for population;

Step 6: carry out retaining via improved adaptive GA-IAGA (i.e. HGA), select, intersect, make a variation, copy operation, adopt the evaluation function based on sequence, test to confirm that it meets 4 constraintss in step 5 to new generation population, specify evolutionary generation d until improved adaptive GA-IAGA reaches or meet following convergence conditions:

$\underset{v\∈N}{\mathrm{\Σ}}|{\mathrm{\η}}_v^d-{\mathrm{\η}}_v^{d-1}|\≤\mathrm{\ϵ}---\left(7\right)$

(1) in formula:

for chromosome, d is evolutionary generation, and ε is convergence criterion;

Step 7: using final for the optimum individual in chromosome as the buffering adjustmentcapacity of unit after optimizing with abandon air quantity, and according to buffering adjustmentcapacity of unit after optimization with abandon air quantity and carry out Real-Time Scheduling.

Wind-powered electricity generation predicated error is accessed corresponding access point as negative load by the present invention, based on the Probabilistic Load Flow method of Latin hypercube, solves the probability distribution of circuit effective power flow, utilizes the reserve capacity of buffering unit to dissolve wind-powered electricity generation predicated error simultaneously.Adopt constraints condition of opportunity, given trend constraint retrains with reserve capacity the confidence level set up, and calculates optimum results eventually through improved adaptive GA-IAGA.

Further, in the present invention, the wind-powered electricity generation predicated error of Wind turbines is considered as stochastic variable, the wind-powered electricity generation predicated error amount of described Wind turbines random distribution adopts laplacian distribution to describe, and the probability density function of corresponding wind-powered electricity generation predicated error is:

$F\left(x\right)=\frac{1}{2}\mathrm{\λexp}(-\mathrm{\λ}|x\left|\right)---\left(8\right)$

In formula, parameter lambda >0, can the probability density function of matching wind-powered electricity generation predicated error by wind energy turbine set a large amount of history wind power output deviation data, thus compares with laplacian distribution and obtain parameter lambda.

Beneficial effect:

Provided by the inventionly a kind ofly consider that the unit of wind-powered electricity generation predicated error is gained merit real-time scheduling method, laplacian distribution is used to describe wind-powered electricity generation predicated error random distribution nature, optimum with to abandon wind minimum for Bi-objective with economy, introduce constraints condition of opportunity, based on the Probabilistic Load Flow method solving system effective power flow probability distribution of Latin Hypercube Sampling, set up and consider that the buffering unit of wind-powered electricity generation predicated error is gained merit optimal allocation model, and adopt Revised genetic algorithum to solve this Chance-Constrained Programming Model.

Specifically comprise the advantage of following several respects:

1. after the electric power system Real-Time Scheduling stage considers wind-powered electricity generation predicated error, can avoid wind-powered electricity generation predict may exist compared with the impact of big error on system cloud gray model, the power-balance of guarantee system and Network Security Constraints, the reasonably operation plan of adjustment System, can also while wind power output deviation of dissolving, wind-powered electricity generation for subsequent period predicts that the change of exerting oneself is ready, receives wind power generation as much as possible;

2. computing system effective power flow distribution in existing real-time scheduling, generally adopts linearization technique, introduces load balancing sensitivity, there is certain error.Probabilistic Load Flow algorithm, by considering the uncertainty of each state variable in Load flow calculation, can obtain the probability distribution of node voltage and Line Flow etc. more accurately.Monte-Carlo Simulation is a kind of Probabilistic Load Flow algorithm the most often adopted, but when sampling larger, amount of calculation is also very huge.The present invention adopts the Probabilistic Load Flow algorithm based on Latin Hypercube Sampling, effectively can improve computational efficiency;

3. owing to there is stochastic variable, deterministic constraints is no longer applicable, and adopt chance constrained programming method establishment buffering unit to gain merit optimal allocation model, the confidence level that given constraint is set up, fully takes into account the random distribution of wind-powered electricity generation predicated error.

In sum, the present invention considers wind-powered electricity generation predicated error in Real-Time Scheduling, the power imbalances that wind power output deviation can be avoided to bring to system and trend out-of-limit, the safe operation of safeguards system.This external system, while wind-powered electricity generation predicated error of dissolving, improves wind-powered electricity generation and receives ability.

Accompanying drawing explanation

To be that the present invention is a kind of consider that the unit of wind-powered electricity generation predicated error is gained merit the algorithm flow chart of real-time scheduling method to Fig. 1;

Fig. 2 is example IEEE24 node system used in embodiment;

Fig. 3 is that unit 22 that in embodiment, Case Simulation obtains is exerted oneself comparison diagram;

Fig. 4 is that unit 24 that in embodiment, Case Simulation obtains is exerted oneself comparison diagram.

Embodiment

Below in conjunction with accompanying drawing, the present invention is further described.

Be illustrated in figure 1 flow chart of the present invention.The present invention implements according to following steps.

Step one: read and treat scheduling power system network data and unit parameter, and the unit output after up-to-date rolling scheduling in electric power system;

Step 2, setting buffering unit optimal allocation model target function of gaining merit are

$\min (\underset{v\∈N_{\mathrm{rt}}}{\mathrm{\Σ}}{a}_v^{\mathrm{rt}}\·\mathrm{\Δ}{P}_v^{\mathrm{rt}}\left(t\right)+\underset{v\∈N_{\mathrm{wind}}}{\mathrm{\Σ}}{\mathrm{\κ}}_v\·\mathrm{\Δ}{P}_v^{\mathrm{loss}}\left(t\right))---\left(1\right)$

(1) in formula:

V represents that unit comprises buffering unit and Wind turbines, N _rtthe set of buffering unit, N _windthe set of Wind turbines, a _rtbuffering unit generated output unit Setup Cost, cushion unit at t adjustment amount of exerting oneself just, wind turbines t just abandon air quantity, κ _vfor abandoning wind weight factor;

Step 3, obtain the wind-powered electricity generation prediction in next cycle and exert oneself and total load information of forecasting, and threshold value P is set _wind; Utilize the wind-powered electricity generation in next cycle to predict to exert oneself the wind power output deducting current period to obtain wind-powered electricity generation gross capability prediction increment, judge that wind-powered electricity generation gross capability predicts whether increment is greater than P _wind; If be more than or equal to P _wind, then next cycle power Systematic selection abandons wind; If be less than P _wind, then next cycle power Systematic selection all receives wind power output;

Step 4: setting with for decision variable, form chromosome to be asked;

Step 5, random generation, just for chromosome, obtain wind-powered electricity generation predicated error sample based on Latin Hypercube Sampling, obtain electric power system effective power flow probability distribution, verify whether it meets following 4 constraintss by Probabilistic Load Flow method;

$\underset{v\∈N_{\mathrm{rt}}}{\mathrm{\Σ}}\mathrm{\Δ}{P}_v^{\mathrm{rt}}\left(t\right)+\underset{v\∈N_{\mathrm{wind}}}{\mathrm{\Σ}}(\mathrm{\Δ}{P}_v^{\mathrm{wind}}\left(t\right)-\mathrm{\Δ}{P}_v^{\mathrm{loss}}\left(t\right))-\mathrm{\Δ}{P}_{\mathrm{load}}=0---\left(2\right)$

$\begin{array}{cc}\mathrm{Pr}\{\underset{\‾}{P_{\mathrm{Tk}}}\≤P_{\mathrm{Tk}}\left({\mathrm{\ϵ}}_v^f\left(t\right)\right)\≤\underset{\‾}{\overset{\‾}{P_{\mathrm{Tk}}}}\}\≥1-{\mathrm{\θ}}_1^{\mathrm{rt}}& \left(4\right)\\ -\mathrm{\Δ}\underset{\‾}{P_v^{\mathrm{rt}}}\≤\mathrm{\Δ}{P}_v^{\mathrm{rt}}\left(t\right)\≤\mathrm{\Δ}{P}_v^{\mathrm{rt}}& \left(5\right)\end{array}$

$\mathrm{Pr}\{-\underset{v\∈N_{\mathrm{rt}}}{\mathrm{\Σ}}\mathrm{\Δ}\underset{\‾}{P_v^{\mathrm{rt}}}\≤\underset{v\∈N_{\mathrm{rt}}}{\mathrm{\Σ}}\mathrm{\Δ}{P}_v^{\mathrm{rt}}\left(t\right)-\underset{v\∈N_{\mathrm{wind}}}{\mathrm{\Σ}}({\mathrm{\ϵ}}_v^f\left(t\right)-\mathrm{\Δ}{P}_v^{\mathrm{loss}}\left(t\right))\≤\underset{v\∈N_{\mathrm{rt}}}{\mathrm{\Σ}}\mathrm{\Δ}\stackrel{\‾}{P_v^{\mathrm{rt}}}\}\≤1-{\mathrm{\θ}}_2^{\mathrm{rt}}---\left(6\right)$

In formula:

(2) in formula: that Wind turbines predicts the increment of exerting oneself, Δ P _loadfor the increment of total load prediction, total load prediction increment deducts current period total load for next cycle total load information of forecasting;

(3) formula is chance constraint, wherein: for the wind-powered electricity generation predicated error amount of Wind turbines, P _tkfor the active power probability distribution of the line section k after t Real-Time Scheduling wind-powered electricity generation predicated error, because load prediction error is much smaller than wind-powered electricity generation predicated error, can think that the variation tendency of super short period load is can by accurately predicting, so disregard the error of load prediction herein for the active power transfer upper limit of line section k, for the active power transfer lower limit of line section k, Pr{} represents probability, for given trend constraint confidence level, and this formula all need be set up section k arbitrary in electrical network;

(4) in formula: for the lower limit of buffering unit output adjustment, and $\mathrm{\Δ}\underset{\‾}{P_v^{\mathrm{rt}}}=\min (P_v^{\mathrm{rt}}(t-1)-P_{v,\min },\mathrm{\Δ}{P}_v^{\mathrm{dn}}),$ Wherein P _{v, min}for buffering unit output lower limit, for cushioning the maximum downward amount (being determined by buffering unit climbing capacity) of unit output, for the upper limit of buffering unit output adjustment, and wherein P _{v, max}for buffering unit v exerts oneself the upper limit, for cushioning the maximum rise amount (being determined by buffering unit climbing capacity) of unit output;

(5) formula is chance constraint, wherein: for given spinning reserve constraint confidence level; Buffering unit needs to stay sufficient reserve capacity and to dissolve wind-powered electricity generation predicated error the power-balance of guarantee system.

If meet institute's Prescribed Properties, then the first of correspondence is put into just for population for chromosome; Otherwise regenerate chromosome and repeat above-mentioned sampling and verification method on new chromosomal basis, being put into by the chromosome of the correspondence meeting constraints just for population, until just reach set point N for the population scale of population;

Step 6: carry out retaining via improved adaptive GA-IAGA, select, intersect, make a variation, copy operation, and adopt the evaluation function based on sequence in operating process, specifically can target function be minimised as permutation function as evaluation function, test to confirm that it meets 4 constraintss in step 5 to new generation population, specify evolutionary generation d until improved adaptive GA-IAGA reaches or meet following convergence conditions:

$\underset{v\∈N}{\mathrm{\Σ}}|{\mathrm{\η}}_v^d-{\mathrm{\η}}_v^{d-1}|\≤\mathrm{\ϵ}---\left(7\right)$

(1) in formula:

for chromosome, d is evolutionary generation, and ε is convergence criterion;

Step 7: using final for the optimum individual in chromosome as the buffering adjustmentcapacity of unit after optimizing with abandon air quantity, and according to buffering adjustmentcapacity of unit after optimization with abandon air quantity and carry out Real-Time Scheduling.

Example:

The present invention adopts IEEE24 node system as example, and the network topology structure of system as shown in Figure 2.Buffering unit is the 21st, 22,23,24,25, No. 26 unit, and the maximum adjustment amount of buffering unit output is 30% of its maximum output.The access point of wind energy turbine set is 1,2,7,15,16, No. 23 node, and the percentage that the wind-powered electricity generation capacity at each access point place accounts for wind energy turbine set gross capability is respectively 15%, 15%, 15%, 15%, 20%, 20%.The wind-powered electricity generation predicated error that each wind energy turbine set shifts to an earlier date 15min all obeys Laplacian probability distribution, and the parameter lambda of corresponding probability density function is 38.22.Use the wind-powered electricity generation database in ERCOT database to produce 96 period wind-powered electricity generation information of forecastings, getting total average annual wind power is 263.9558MW.

After MATLAB emulation, obtain 96 periods Real-Time Scheduling buffering unit output adjustment situation in a day.In emulation, confidence level is set to ensure that power flow stability and the power-balance of system.Fig. 3 is the change of exerting oneself considering to cushion before and after wind-powered electricity generation predicated error unit 22, and Fig. 4 is the change of exerting oneself considering to cushion before and after wind-powered electricity generation predicated error unit 24.As can be seen from Fig. 3 and Fig. 4, consider that before and after wind-powered electricity generation predicated error, the general trend of unit output is consistent.And No. 22 unit output scopes are 84 ~ 172MW, No. 24 unit output scopes are 160 ~ 340MW, all do not exceed the bound of corresponding unit output, and what buffering unit was described arranges the optimization aim that can complete Real-Time Scheduling well.But owing to considering wind-powered electricity generation predicated error, exerting oneself of unit of buffering also there occurs change, and unit output leaves more bidirectional modulation nargin.

In most cases, it is 0MW that the wind energy turbine set in Real-Time Scheduling abandons air quantity.In order to verify the feasibility proposing real-time scheduling method, suitably reduce threshold value P _windvalue, and do not consider that wind-powered electricity generation predicated error is abandoned wind and contrasted, abandoning air quantity when not considered wind-powered electricity generation predicated error is 42.5316MW, and abandoning air quantity after considering wind-powered electricity generation predicated error is 26.3336MW.Can find out, after considering wind-powered electricity generation predicated error, abandon air quantity and significantly decline.This is because buffering unit is dissolving wind power output deviation while, also for the wind-powered electricity generation of subsequent period predicts that the change of exerting oneself is ready, adjusts the distribution of buffering unit output in advance, receives wind power generation as much as possible, and system receives wind-powered electricity generation change capacity to increase.

The above is only the preferred embodiment of the present invention; be noted that for those skilled in the art; under the premise without departing from the principles of the invention, can also make some improvements and modifications, these improvements and modifications also should be considered as protection scope of the present invention.

Claims (2)

1. consider that the unit of wind-powered electricity generation predicated error is gained merit a real-time scheduling method, it is characterized in that: comprise the following steps that order performs:

Step one, read and wait to dispatch the network data of electric power system and unit parameter, and the unit output after up-to-date rolling scheduling in electric power system;

Step 2, setting buffering unit optimal allocation model target function of gaining merit are

$\min (\underset{v\∈N_{\mathrm{rt}}}{\mathrm{\Σ}}{a}_v^{\mathrm{rt}}\·\mathrm{\Δ}{P}_v^{\mathrm{rt}}\left(t\right)+\underset{v\∈N_{\mathrm{wind}}}{\mathrm{\Σ}}{\mathrm{\κ}}_v\·\mathrm{\Δ}{P}_v^{\mathrm{loss}}\left(t\right))---\left(1\right)$

(1) in formula:

V represents that unit comprises buffering unit and Wind turbines, N _rtthe set of buffering unit, N _windthe set of Wind turbines, buffering unit generated output unit Setup Cost, cushion unit at t adjustment amount of exerting oneself just, wind turbines t just abandon air quantity, κ _vfor abandoning wind weight factor;

Step 3, obtain the wind-powered electricity generation prediction in next cycle and exert oneself and total load information of forecasting, and threshold value P is set _wind; Utilize the wind-powered electricity generation in next cycle to predict to exert oneself the wind power output deducting current period to obtain wind-powered electricity generation gross capability prediction increment, judge that wind-powered electricity generation gross capability predicts whether increment is greater than P _wind; If be more than or equal to P _wind, then next cycle power Systematic selection abandons wind; If be less than P _wind, then next cycle power Systematic selection all receives wind power output;

Step 4, setting with for decision variable, form chromosome to be asked;

Step 5, random generation, just for chromosome, obtain wind-powered electricity generation predicated error sample based on Latin Hypercube Sampling, obtain electric power system effective power flow probability distribution, verify whether it meets following 4 constraintss by Probabilistic Load Flow method;

$\underset{v\∈N_{\mathrm{rt}}}{\mathrm{\Σ}}\mathrm{\Δ}{P}_v^{\mathrm{rt}}\left(t\right)+\underset{v\∈N_{\mathrm{wind}}}{\mathrm{\Σ}}(\mathrm{\Δ}{P}_v^{\mathrm{wind}}\left(t\right)-\mathrm{\Δ}{P}_v^{\mathrm{loss}}\left(t\right))-\mathrm{\Δ}{P}_{\mathrm{load}}=0---\left(2\right)$

$\mathrm{Pr}\{\underset{\‾}{P_{\mathrm{Tk}}}\≤P_{\mathrm{Tk}}\left({\mathrm{\ϵ}}_v^f\left(t\right)\right)\≤\underset{\‾}{\stackrel{\‾}{P_{\mathrm{Tk}}}}\}\≥1-{\mathrm{\θ}}_1^{\mathrm{rt}}---\left(4\right)$

$-\mathrm{\Δ}\underset{\‾}{P_v^{\mathrm{rt}}}\≤\mathrm{\Δ}{P}_v^{\mathrm{rt}}\left(t\right)\≤\mathrm{\Δ}{P}_v^{\mathrm{rt}}---\left(5\right)$

$\mathrm{Pr}\{-\underset{v\∈N_{\mathrm{rt}}}{\mathrm{\Σ}}\mathrm{\Δ}\underset{\‾}{P_v^{\mathrm{rt}}}\≤\underset{v\∈N_{\mathrm{rt}}}{\mathrm{\Σ}}\mathrm{\Δ}{P}_v^{\mathrm{rt}}\left(t\right)-\underset{v\∈N_{\mathrm{wind}}}{\mathrm{\Σ}}({\mathrm{\ϵ}}_v^f\left(t\right)-\mathrm{\Δ}{P}_v^{\mathrm{loss}}\left(t\right))\≤\underset{v\∈N_{\mathrm{rt}}}{\mathrm{\Σ}}\mathrm{\Δ}\stackrel{\‾}{P_v^{\mathrm{rt}}}\}\≤1-{\mathrm{\θ}}_2^{\mathrm{rt}}---\left(6\right)$

In formula:

(2) in formula: that Wind turbines predicts the increment of exerting oneself, Δ P _loadfor the increment of total load prediction, total load prediction increment deducts current period total load for next cycle total load information of forecasting;

(3) in formula: for the wind-powered electricity generation predicated error amount of Wind turbines, P _tkfor the active power probability distribution of the line section k after t Real-Time Scheduling wind-powered electricity generation predicated error, for the active power transfer upper limit of line section k, for the active power transfer lower limit of line section k, Pr{} represents probability, for given trend constraint confidence level, and this formula all need be set up section k arbitrary in electrical network;

(4) in formula: for the lower limit of buffering unit output adjustment, and $\mathrm{\Δ}\underset{\‾}{P_v^{\mathrm{rt}}}=\min (P_v^{\mathrm{rt}}(t-1)-P_{v,\min },\mathrm{\Δ}{P}_v^{\mathrm{dn}}),$ Wherein P _{v, min}for buffering unit output lower limit, for cushioning the maximum downward amount of unit output, for the upper limit of buffering unit output adjustment, and $\mathrm{\Δ}\stackrel{\‾}{P_v^{\mathrm{rt}}}=\min (P_{v,\max }-P_v^{\mathrm{rt}}(t-1),\mathrm{\Δ}{P}_v^{\mathrm{up}}),$ Wherein P _{v, max}for the buffering unit output upper limit, for cushioning the maximum rise amount of unit output;

(5) in formula: for given spinning reserve constraint confidence level;

If meet institute's Prescribed Properties, then the first of correspondence is put into just for population for chromosome; Otherwise regenerate chromosome and repeat above-mentioned sampling and verification method on new chromosomal basis, the chromosome of the correspondence meeting constraints being put into just for population, until reach population scale N;

Step 6: carry out retaining via improved adaptive GA-IAGA, select, intersect, make a variation, copy operation, adopt the evaluation function based on sequence, test to confirm that it meets 4 constraintss in step 5 to new generation population, specify evolutionary generation d until improved adaptive GA-IAGA reaches or meet following convergence conditions:

$\underset{v\∈N}{\mathrm{\Σ}}|{\mathrm{\η}}_v^d-{\mathrm{\η}}_v^{d-1}|\≤\mathrm{\ϵ}---\left(7\right)$

(1) in formula:

for chromosome, d is evolutionary generation, and ε is convergence criterion;

Step 7: using final for the optimum individual in chromosome as the buffering adjustmentcapacity of unit after optimizing with abandon air quantity, and according to buffering adjustmentcapacity of unit after optimization with abandon air quantity and carry out Real-Time Scheduling.

2. according to claim 1ly a kind ofly consider that the unit of wind-powered electricity generation predicated error is gained merit real-time scheduling method, it is characterized in that: the wind-powered electricity generation predicated error amount of Wind turbines random distribution adopts laplacian distribution to describe.