CN105956713A - New energy annual/monthly electric quantity plan making method - Google Patents

Abstract

The present invention provides a new energy annual/monthly electric quantity plan making method considering randomness. The method comprises the steps of generating a plurality of new energy output sequences randomly according to a new energy annual/monthly electric quantity prediction result; using a timing simulation calculation method to calculate the new energy consumed electric quantity on different new energy output curve conditions; obtaining an expected value of the calculated new energy consumed electric quantity, and obtaining the final new energy consumed electric quantity, thereby forming a new energy annual/monthly electric quantity plan. The method considers the influence of the new energy long-time scale randomness and the electric power system operation constraint on the new energy electric quantity plan, and enables the accuracy of the new energy annual/monthly electric quantity plan to be improved effectively.

Description

A kind of new forms of energy years months electricity ways to draw up the plan

Technical field

The present invention relates to electricity plan, be specifically related to a kind of new forms of energy years months coulant meter considering randomness and draw Determine method.

Background technology

Electricity plan is one of important process of Power System Planning management.In recent years, in the face of energy crisis, finance Climate crisis is the most clearly recognized by crisis and the mankind, and in global range, the situation of extra normal development occur in new forms of energy, The investment of new forms of energy is increased substantially by various countries, and new forms of energy production capacity the most drastically expands.Quick increasing along with new forms of energy installed capacity Long so that some areas new forms of energy are difficult to fully dissolve, in the urgent need to a kind of new forms of energy years months considering new forms of energy limited situation Electricity ways to draw up the plan.

The present invention proposes a kind of new forms of energy years months electricity ways to draw up the plan considering randomness, meets generation of electricity by new energy Feature；Consider generation of electricity by new energy limited situation, make calculated new forms of energy years months electricity plan more conform to new forms of energy The actual demand of generating.

Summary of the invention

The present invention proposes a kind of new forms of energy years months electricity ways to draw up the plan considering randomness；The method is according to new Energy years months power quantity predicting result, a plurality of new forms of energy of stochastic generation are exerted oneself sequence；Use the method that time stimulatiom calculates, calculate In the case of different new forms of energy power curves, new forms of energy are dissolved electricity；Electricity of dissolving calculated new forms of energy takes expected value, Dissolve electricity to final new forms of energy, thus form new forms of energy years months electricity plan.The method considers the long-time chi of new forms of energy Degree randomness and the Operation of Electric Systems constraint impact on new forms of energy electricity plan, be effectively increased new forms of energy years months coulant meter The accuracy drawn.

A kind of new forms of energy years months electricity ways to draw up the plan, comprises the steps:

Step 1: generate new forms of energy years months prediction electricity；

Step 2: a plurality of new forms of energy of stochastic generation are exerted oneself sequence；

Step 3: calculate different new forms of energy new forms of energy under sequence of exerting oneself and dissolve electricity；

Step 4: electricity of dissolving the new forms of energy obtained takes expected value.

Described step 1 uses kalman filter method and autoregressive moving average method to predict medium-term and long-term meteorological data, point Do not calculate the prediction electricity of wind-powered electricity generation, photovoltaic.

It is as follows that the pre-quantometer of described wind-powered electricity generation calculates method:

The calculating Weibull distribution parameters based on statistical property of wind-powered electricity generation prediction electricity, the wind frequency curve of wind-powered electricity generation prediction electricity It is shown below:

$P\left(v\right)=\frac{k}{c}{\left(\frac{v}{c}\right)}^{k-1}\exp \[-{\left(\frac{v}{c}\right)}^k\]---\left(1\right)$

In formula, v is wind speed, k and c is Weibull distribution parameters；K is form parameter, determines the change model of mean wind speed Enclose；C is scale parameter, determines the peak value size of wind frequency curve.

For continuous, the air monitoring data of high frequency, use mean wind speedWith standard deviation S_vEstimation Weibull distribution ginseng Number；

$\stackrel{\‾}{v}=\frac{1}{n}\underset{i=1}{\overset{n}{\Σ}}{V}_i---\left(2\right)$

$S_v=\sqrt{\frac{1}{n}\underset{i=1}{\overset{n}{\Σ}}{(V_i-\stackrel{\‾}{v})}^2}---\left(3\right)$

$k={\left(\frac{S_v}{\stackrel{\‾}{v}}\right)}^{-1.086}---\left(4\right)$

$c=\frac{\stackrel{\‾}{v}}{\mathrm{\Γ}(1+1/k)}---\left(5\right)$

In formula, V_iFor Wind observation sequence, n is the number of wind series in calculation interval, and Γ (1+1/k) is gamma letter Number；Wind energy turbine set is at wind speed v_iUnder level, generated energy is shown below:

W_j=f_j(v_i)·P_j(v_i)·S·δ (6)

In formula, f_j() is the actual power curve of jth wind energy turbine set, can be used with actual power data acquisition by historical wind speed data Distribution-free regression procedure matching obtains, P_j(v_i) it is at wind speed v_iThe wind frequency curve of jth wind energy turbine set under level, S is that sample always holds Amount, δ is that wind speed uses frequency；Years months wind-powered electricity generation prediction electricity is by obtaining the electricity under all velocity wind levels is cumulative.

It is as follows that the pre-quantometer of described photovoltaic calculates method:

With the monthly average irradiance of the sun for photovoltaic electricity Prediction Parameters, the monthly average of t photovoltaic plant region Irradiation intensity is E_t, then t photovoltaic plant exert oneself for:

P_t=a E_t+b (7)

In formula, P_tFor exerting oneself of t photovoltaic plant, a is coefficient of first order, and b is constant term；Parameter a, b, use a young waiter in a wineshop or an inn Multiplication carries out identification；Photovoltaic prediction electricity is always adding of exerting oneself of all photovoltaic plants.

Step 1 is predicted the total electricity result obtained by described step 2, uses stochastic modeling method, and stochastic generation is a plurality of newly The energy is exerted oneself sequence.

The a plurality of new forms of energy that step 2 is generated by described step 3 are exerted oneself sequence, carry out sequential and produce analog simulation and calculate, meter Calculate different new forms of energy new forms of energy under sequence of exerting oneself to dissolve electricity；

Sequential produce analog simulation calculate consider the wind-powered electricity generation/photovoltaic of electrical network exert oneself sequence characteristic, load temporal characteristics, Peak load regulation characteristic, electrical network send ability, set up provincial power network time stimulatiom model, optimize the power balance of the whole network by the period.

Described step 4 electricity of dissolving the new forms of energy that step 3 calculates takes expected value, and as new forms of energy years months electricity Gauge is drawn.

With immediate prior art ratio, technical scheme has following excellent effect and is:

1, the present invention take into account randomness that generation of electricity by new energy exerts oneself and new forms of energy are rationed the power supply to new forms of energy electricity plan Impact, be effectively increased the reliability of new forms of energy years months electricity plan, effectively reduce generation of electricity by new energy and be obstructed electricity.

2, the present invention carries out time stimulatiom by constructing multiple new forms of energy time series of exerting oneself, and is made by the expected value of result For final result, it is to avoid the shortcoming that single calculation randomness is big.

Accompanying drawing explanation

Fig. 1 is the calculation flow chart of a kind of new forms of energy years months electricity ways to draw up the plan considering randomness of the present invention.

Detailed description of the invention

The present invention relates to a kind of new forms of energy years months electricity ways to draw up the plan considering randomness, the method considers new energy The randomness exerted oneself in source and generation of electricity by new energy limited situation, can improve the reliability of new forms of energy years months electricity plan, below knot Conjunction accompanying drawing is embodied as flow process to the present invention and is described.

Step 1 is to generate new forms of energy years months prediction electricity；Use kalman filter method and autoregressive moving average method Predicting medium-term and long-term meteorological data, and then calculate the prediction electricity of wind-powered electricity generation, photovoltaic respectively, computational methods are as follows:

Wind-powered electricity generation prediction electricity Weibull distribution parameters based on statistical property, Two-parameter Weibull Distribution is a kind of unimodal Two-parameter distribution function bunch, its probability density function (wind frequency curve) is:

$P\left(v\right)=\frac{k}{c}{\left(\frac{v}{c}\right)}^{k-1}\exp \[-{\left(\frac{v}{c}\right)}^k\]---\left(1\right)$

In formula, v is wind speed, k and c is Weibull distribution parameters, and k is form parameter, determines the change model of mean wind speed Enclosing, c is scale parameter, determines the peak value size of wind frequency curve.For continuous, the air monitoring data of high frequency, use average wind SpeedWith standard deviation S_vEstimation Weibull distribution parameters.

$\stackrel{\‾}{v}=\frac{1}{n}\underset{i=1}{\overset{n}{\Σ}}{V}_i---\left(2\right)$

$S_v=\sqrt{\frac{1}{n}\underset{i=1}{\overset{n}{\Σ}}{(V_i-\stackrel{\‾}{v})}^2}---\left(3\right)$

$k={\left(\frac{S_v}{\stackrel{\‾}{v}}\right)}^{-1.086}---\left(4\right)$

$c=\frac{\stackrel{\‾}{v}}{\mathrm{\Γ}(1+1/k)}---\left(5\right)$

In formula, V_iFor Wind observation sequence, n is the number of wind series in calculation interval, and Γ (1+1/k) is gamma letter Number, can look into gamma function table and try to achieve.Wind energy turbine set is at wind speed v_iUnder level can generated energy be

W_j=f_j(v_i)·P_j(v_i)·S·δ (6)

In formula, f_j() is the actual power curve of jth wind energy turbine set, can be by historical wind speed data and actual power data Distribution-free regression procedure matching is used to obtain, P_j(v_i) it is at wind speed v_iThe wind frequency curve of jth wind energy turbine set under level, S is that sample is total Capacity, δ is that wind speed uses frequency.Just monthly, annual prediction electricity is can get to electricity under all velocity wind levels is cumulative.

Photovoltaic resources is directly determined by solar irradiance, is mainly affected by latitude and height above sea level, areal year spoke Illumination has certain stability.Photovoltaic electricity Prediction Parameters is monthly average irradiance, and irradiance and photovoltaic plant are exerted oneself existence Positive linear relationships, it is assumed that the irradiation intensity of t photovoltaic plant region is E_t, then t photovoltaic plant exert oneself for:

P_t=a E_t+b (7)

In formula, a is coefficient of first order, and b is constant term.Photovoltaic power generation quantity monthly, annual be all photovoltaic plants exert oneself total Add.For parameter a, b, method of least square can be used to carry out identification.

Step 2 is the new forms of energy annual prediction electricity generated according to step 1, and a plurality of new forms of energy of stochastic generation are exerted oneself sequence； New forms of energy Series Modeling of exerting oneself is total electricity result resources obtained, and uses stochastic modeling method, stochastic generation wind-powered electricity generation/ Photovoltaic generation is exerted oneself sequence；Wind-powered electricity generation and photovoltaic modeling method are described as follows:

Wind power output, based on history wind-powered electricity generation sequence, is regarded as the ripple that First air process causes by wind-powered electricity generation time series stochastic modeling Dynamic, intensity based on wind process is divided into great fluctuation process, middle fluctuation, minor swing and low go out fluctuation, and use visual from group Knit mapping (Self-Organizing Map, SOM) clustering algorithm and carry out identification of fluctuating, and statistics obtains what all kinds of fluctuation occurred Number of times, fluctuation height, persistent period, the probability of a kind of fluctuation to another kind of fluctuation transfer；It is then based on random sequential sampling, To the wind power output sequence meeting these indexs.

Photovoltaic sequence stochastic modeling method of exerting oneself is exerted oneself sequence based on history photovoltaic, is broken down into headroom definitiveness part With weather characteristics uncertainty part；For weather characteristics uncertainty part, synoptic process is divided into fine day, broken sky, the moon My god, sudden change weather four type, same use random sequential sampling techniques to obtain photovoltaic to exert oneself sequence.

Step 3 is that a plurality of new forms of energy generating step 2 are exerted oneself sequence, carries out sequential and produces analog simulation and calculate, calculates Different new forms of energy new forms of energy under sequence of exerting oneself are dissolved electricity；Sequential produce analog simulation calculate consider the wind-powered electricity generation of electrical network/ Photovoltaic time series characteristic, load temporal characteristics, peak load regulation characteristic, electrical network send the factors such as ability, when setting up provincial power network Sequence phantom, optimizes the power balance of the whole network by the period.Model constraints and object function are as follows:

(1) set optimization power constraint

$X_j^t\·P_{j,min}\≤P_j\left(t\right)\≤P_{j,max}\·X_j^t---\left(8\right)$

(2) minimum start and stop time-constrain

$Y_j^t+\underset{i=1}{\overset{k_{on}}{\Σ}}{Z}_j^{t+i}\≤1---\left(9\right)$

$Z_j^t+\underset{i=1}{\overset{k_{off}}{\Σ}}{Y}_j^{t+i}\≤1---\left(10\right)$

(3) heat supply phase thermal power plant unit units limits

$P_{j,BY}^t=C_j^b\·H_j^t---\left(11\right)$

$H_j^t\·C_j^b\≤P_{j,CQ}^t\≤P_{j,max}-H_j^t\·C_j^v---\left(12\right)$

(4) start and stop logic state constraint

$\left\{\begin{array}{c}{X}_j^t-X_j^{t-1}-Y_j^t+Z_j^t=0\\ -X_j^t-X_j^{t-1}+Y_j^t\≤0\\ X_j^t+X_j^{t-1}+Y_j^t\≤2\\ -X_j^t-X_j^{t-1}+Z_j^t\≤0\\ X_j^i+X_j^{t-1}+Z_j^i\≤2\end{array}\right.---\left(13\right)$

(5) unit climbing rate constraint

$P_j^{t+1}-P_j^t\≤{\mathrm{\ΔP}}_{j,up}---\left(14\right)$

$P_j^t-P_j^{t+1}\≤{\mathrm{\ΔP}}_{j,down}---\left(15\right)$

(6) spinning reserve constraint

$\begin{array}{c}-\underset{j=1}{\overset{J}{\Σ}}{P}_{j,\max }\·X_j^t\≤-P_l^t-\mathrm{Pr}e\\ \underset{j=1}{\overset{J}{\Σ}}{P}_{j,\min }\·X_j^t\≤P_l^t-Nre\end{array}---\left(16\right)$

(7) interregional line transmission capacity-constrained

$L_{n,nn}^{\min }\≤L_{n,nn}^t\≤L_{n,nn}^{\max }---\left(17\right)$

(8) new forms of energy power constraint

$0\≤P_{w,n}\left(t\right)\≤P_{w,n}^{*}\left(t\right)---\left(18\right)$

(9) object function

$max\underset{t=1}{\overset{T}{\Σ}}\underset{n=1}{\overset{N}{\Σ}}{P}_{w,n}\left(t\right)---\left(19\right)$

In formula, P_j,max, P_j,minIt is respectively the exert oneself upper limit and the lower limit of exerting oneself of jth platform unit.For operating states of the units, " 1 " represents operation, and " 0 " represents shutdown.It is respectively and represents that unit j starts at period t, the binary system of stopped status becomes Amount,Represent that unit starts for " 1 ", for " 0 " expression unit not at starting state,Represent that unit is shut down for " 1 ", Represent that unit is not in stopped status for " 0 "；k_onThe machine time is opened for unit minimum；k_offFor unit minimum downtime；That reflects Minimum opens the time span of machine or shutdown, and different types of Unit Commitment machine time parameter is different.Exert oneself greatly for back pressure unit Little；Exert oneself size for extraction steam unit；For t period load of heat；For thermal power plant unit coupled thermomechanics coefficient.It is respectively creep speed and lower creep speed in the maximum of unit j.Pre and Nre be respectively positive rotation standby and Negative spinning reserve.For the tie-line power transmission upper limit between t region n and region nn,For t region n and district Tie-line power transmission lower limit between the nn of territory.Exert oneself for new forms of energy theory.

Step 4 is that the new forms of energy electricity of dissolving calculating step 3 takes expected value, and as new forms of energy years months electricity Plan；Assume xth_iThe bar new forms of energy calculated new forms of energy of sequence electricity of dissolving of exerting oneself is F (x_i), then calculate N_sSecondary new energy Dissolve electricity expected value in sourceFormula (20), (21) are seen respectively with convergence criterion β computational methods.

$\hat{E}\left(F\right)=\frac{1}{N_S}\underset{i=1}{\overset{N_S}{\Σ}}F\left(x_i\right)---\left(20\right)$

$\mathrm{\β}=\frac{\sqrt{\underset{i=1}{\overset{N_S}{\Σ}}{\[F\left(x_i\right)-\hat{E}\left(F\right)\]}^2}}{N_S\hat{E}\left(F\right)}---\left(21\right)$

When convergence criterion β meets plan demand, using the new forms of energy years months charge value of corresponding simulation result as newly Energy electricity on days plan.Wherein convergence criterion β typically takes 0.5%, can revise according to the actual requirements.

Finally should be noted that: above example is merely to illustrate technical scheme rather than to its protection domain Restriction, although being described in detail the application with reference to above-described embodiment, those of ordinary skill in the field should Understand: those skilled in the art read the application after still can to application detailed description of the invention carry out all changes, amendment or Person's equivalent, but these changes, amendment or equivalent, all within the claims that application is awaited the reply.

Claims (9)

1. a new forms of energy years months electricity ways to draw up the plan, it is characterised in that described method comprises the steps:

Step 1: generate new forms of energy years months prediction electricity；

Step 2: a plurality of new forms of energy of stochastic generation are exerted oneself sequence；

Step 3: calculate different new forms of energy new forms of energy under sequence of exerting oneself and dissolve electricity；

Step 4: electricity of dissolving the new forms of energy generated takes expected value.

2. new forms of energy years months electricity ways to draw up the plan as claimed in claim 1, it is characterised in that described step 1 karr Graceful filtering method and autoregressive moving average method predict medium-term and long-term meteorological data, calculate the prediction electricity of wind-powered electricity generation, photovoltaic respectively.

3. new forms of energy years months electricity ways to draw up the plan as claimed in claim 2, it is characterised in that described wind-powered electricity generation prediction electricity Amount computational methods are as follows:

The wind frequency curve of wind-powered electricity generation prediction electricity is shown below:

$P\left(v\right)=\frac{k}{c}{\left(\frac{v}{c}\right)}^{k-1}\exp \[-{\left(\frac{v}{c}\right)}^k\]---\left(1\right)$

In formula, v is wind speed, k and c is Weibull distribution parameters；K: determine the form parameter of the excursion of mean wind speed；C: certainly The scale parameter of the peak value size of subduing the wind syndrome frequency curve.

4. new forms of energy years months electricity ways to draw up the plan as claimed in claim 3, it is characterised in that for continuous, high frequency Air monitoring data, with the mean wind speed shown in following formula (2)With standard deviation S shown in following formula (3)_vEstimation following formula (4) and (5) Shown Weibull distribution parameters k and c；

$\stackrel{\‾}{v}=\frac{1}{n}\underset{i=1}{\overset{n}{\Σ}}{V}_i---\left(2\right)$

$S_v=\sqrt{\frac{1}{n}\underset{i=1}{\overset{n}{\Σ}}{(V_i-\stackrel{\‾}{v})}^2}---\left(3\right)$

$k={\left(\frac{S_v}{\stackrel{\‾}{v}}\right)}^{-1.086}---\left(4\right)$

$c=\frac{\stackrel{\‾}{v}}{\mathrm{\Γ}(1+1/k)}---\left(5\right)$

In formula, V_iFor Wind observation sequence, n is the number of wind series in calculation interval, and Γ (1+1/k) is gamma function；Wind-powered electricity generation Field is at wind speed v_iGenerated energy under level is shown below:

W_j=f_j(v_i)·P_j(v_i)·S·δ (6)

In formula, f_j() is the actual power curve of jth wind energy turbine set, can be used with actual power data acquisition by historical wind speed data Distribution-free regression procedure matching obtains, P_j(v_i) it is at wind speed v_iThe wind frequency curve of jth wind energy turbine set under level, S is that sample always holds Amount, δ is that wind speed uses frequency；Years months wind-powered electricity generation prediction electricity is by obtaining the electricity under all velocity wind levels is cumulative.

5. new forms of energy years months electricity ways to draw up the plan as claimed in claim 2, it is characterised in that described photovoltaic prediction electricity Amount computational methods are as follows:

With the monthly average irradiance of the sun for photovoltaic electricity Prediction Parameters, t photovoltaic plant exert oneself as shown in following formula (7):

P_t=a E_t+b (7)

In formula, P_t: exerting oneself of t photovoltaic plant, E_t: the monthly average irradiation intensity of t photovoltaic plant region, a: one Level number, b: constant term, a and b, use method of least square to carry out identification；Photovoltaic prediction electricity is that all photovoltaic plants are exerted oneself Always add.

6. new forms of energy years months electricity ways to draw up the plan as claimed in claim 1, it is characterised in that described step 2 includes: Predicting, according to step 1, total electricity result of obtaining, with stochastic modeling method, a plurality of new forms of energy of stochastic generation are exerted oneself sequence.

7. new forms of energy years months electricity ways to draw up the plan as claimed in claim 1, it is characterised in that described step 3 includes: Exert oneself sequence according to a plurality of new forms of energy that step 2 generates, carry out sequential and produce analog simulation and calculate, obtain different new forms of energy and exert oneself New forms of energy under sequence are dissolved electricity.

8. new forms of energy years months electricity ways to draw up the plan as claimed in claim 7, it is characterised in that described sequential produces mould Intend simulation calculation to include: exert oneself sequence characteristic, load temporal characteristics, peak load regulation characteristic, electrical network according to the wind-powered electricity generation/photovoltaic of electrical network Send ability, set up provincial power network time stimulatiom model, optimize the power balance of the whole network by the period.

9. new forms of energy years months electricity ways to draw up the plan as claimed in claim 1, it is characterised in that described step 4 includes: Expected value is asked for according to the new forms of energy electricity of dissolving that step 3 calculates, and as new forms of energy years months electricity plan.