CN105680458A - Photovoltaic power station smooth output method - Google Patents

Abstract

The invention provides a photovoltaic power station smooth output method. The photovoltaic power station smooth output method comprises the steps of obtaining a data set formed by photovoltaic power station power predicted values; performing fitting on the data set through a polynomial fitting algorithm to obtain an ideal smooth output formula; calculating an ideal smooth output value according to the ideal smooth output formula; and determining an actual smooth output value according to the magnitude relations between the ideal smooth output value and the photovoltaic power station power predicted values. According to the photovoltaic power station smooth output method, the photovoltaic power station power predicted values in the overall planned output section are subjected to polynomial fitting; the finally obtained actual smooth output curve is not delayed; and compared with the method for stabilizing photovoltaic power station power fluctuation through a first-order low-pass filtering method, the photovoltaic power station smooth output method provided by the invention can achieve a more optimized smooth output effect.

Description

Smoothly exert oneself method in a kind of photovoltaic electric station

Technical field

The present invention relates to technical field of electric power, specifically, it relates to smoothly exert oneself method in a kind of photovoltaic electric station.

Background technology

Photovoltaic generation is that the more ripe renewable and clean energy resource that utilizes carries out the technology generated electricity at present, but there is randomness and fluctuation due to solar energy resources, brings a series of impacts to electrical network. The original trend distribution of the fluctuation degree direct influence electrical network of power, when the rate of permeation of photovoltaic generation is in higher level, fluctuation and randomness bring enormous impact can to the operation of electrical network. Fig. 1 is the system schematic at photovoltaic electric station, and multiple stage photovoltaic generation unit is connected on same bar bus jointly, then by the access power system of photovoltaic electric station owner's transformer. In order to guarantee to inject the electric energy held stationary of electrical network, it is necessary to the fluctuation of photovoltaic generation power is stabilized, to reduce photovoltaic generation power fluctuation to the impact of electrical network.

The people such as the aerospace of Tohoku Electric Power university, Yan Gangui utilize first-order low-pass ripple algorithm to realize wind power and fluctuate the control strategy stabilized. High frequency component in wind energy turbine set operation is mainly carried out filtering by this control strategy, reduce the velocity of variation of wind power, for power system provides comparatively stable power stage, accumulator system is then the amplitude being changed output rating by its discharge and recharge, makes the electric energy of injection electrical network more steady.

Owing to sun power and wind energy are all the renewable energy sources with randomness and fluctuation, in technical field of electric power, this kind utilizes first-order low-pass ripple algorithm to carry out the wind power control strategy stabilized of fluctuation and can also be applied to during photovoltaic electric station fluctuation of power stabilizes, namely utilize first-order low-pass ripple algorithm to be stabilized by photovoltaic electric station fluctuation of power.

But find that the output smoothing curve utilizing first-order low-pass ripple algorithm to obtain exists certain time-lag action in actual applications, as shown in Figure 2, thinner line is that wind-powered electricity generation goes out separately force curve, thicker line is that the wind storing cogeneration utilizing this kind of control strategy to obtain smoothly goes out force curve, as ise apparent from FIG. 2, wind storing cogeneration smoothly goes out force curve and lags behind wind-powered electricity generation and go out separately force curve. This is because this kind utilize first-order low-pass ripple algorithm realize the wind power control strategy stabilized of fluctuation be adopt this sampled value and last time filtering output value carry out weighting and obtain this filtering output value, concrete formula is:

Y (n)=α X (n)+(1-α) Y (n-1)

In above formula, α is filter factor;X (n) is this sampled value; Y (n-1) is filtering output value last time; Y (n) is this filtering output value.

Visible, this kind utilizes first-order low-pass ripple algorithm to realize wind power and fluctuates the control strategy also Shortcomings part stabilized.

With reason, when utilizing first-order low-pass ripple algorithm to be stabilized by photovoltaic electric station fluctuation of power, also unavoidably there will be obtain smoothly go out the delayed problem of force curve time delay.

Summary of the invention

The main purpose of the embodiment of the present invention is to provide a kind of photovoltaic electric station smoothly to exert oneself method, with solve prior art utilize first-order low-pass ripple algorithm to stabilize photovoltaic electric station fluctuation of power obtains smoothly goes out the problem that force curve exists time delay phenomenon.

In order to realize above-mentioned purpose, the embodiment of the present invention provides a kind of photovoltaic electric station smoothly to exert oneself method, comprising:

Steps A, obtains the data acquisition of predicting power of photovoltaic plant value composition;

Step B, utilizes polynomial fitting algorithm that described data acquisition is carried out matching, obtains desired smooth and exerts oneself formula;

Step C, goes out force value according to described desired smooth formulae discovery desired smooth of exerting oneself;

Step D, goes out force value and the magnitude relationship of described predicting power of photovoltaic plant value, it is determined that reality smoothly goes out force value according to described desired smooth;

Described steps A is specially the data acquisition P obtaining predicting power of photovoltaic plant value composition:

P={ (p_i,t_i) | i=1,2..., m};

Wherein, P is data acquisition, p_iFor predicting power of photovoltaic plant value, m is the sample number of described data acquisition and m is natural number, and i is sample sequence number, t_iFor p_iThe corresponding time;

Described step B specifically comprises:

Step B1, according to predicting power of photovoltaic plant value p in described data acquisition P_iFluctuation tendency, it is determined that described desired smooth is exerted oneself the rank number n of formula, and wherein n is natural number;

Step B2, matching has the polynomial expression of described rank number n:

a_nt_i ⁿ+a_n-1t_i ^n-1+…+a₁t_i+a₀;

Wherein, a₀～a_nFor multinomial coefficient;

Step B3, calculates described polynomial expression a_nt_i ⁿ+a_n-1t_i ^n-1+…+a₁t_i+a₀With described predicting power of photovoltaic plant value p_iSquared difference and Err:

$\mathrm{Err}=\underset{i=0}{\overset{m}{\mathrm{\Σ}}}{(a_n{t_i}^n+a_{n-1}{t_i}^{n-1}+...+a_1{t}_i+a_0-p_i)}^2;$

Step B4, when utilizing method of least squares to calculate described squared difference and Err for minimum value, multinomial coefficient a₀～a_nCorresponding concrete value α₀～α_n;

Step B5, utilizes described concrete value α₀～α_nBuild desired smooth to exert oneself formula X (t):

X (t)=α_ntⁿ+α_n-1t^n-1+…+α₁t+α₀;

Wherein, t is the time;

Described step C is specially:

Calculate and work as t=t_iTime, described desired smooth is exerted oneself the value X (t of formula X (t)_i):

X(t_i)=α_nt_i ⁿ+α_n-1t_i ^n-1+…+α₁t_i+α₀

Wherein, X (t_i) go out force value for desired smooth;

Described step D specifically comprises:

When described desired smooth go out force value be less than or equal described predicting power of photovoltaic plant value time, it is determined that reality smoothly goes out force value and equals described desired smooth and go out force value;

When described desired smooth go out force value be greater than described predicting power of photovoltaic plant value time, it is determined that reality smoothly goes out force value and equals described predicting power of photovoltaic plant value.

By means of technique scheme, by whole plan being exerted oneself, interval predicting power of photovoltaic plant value carries out polynomial fitting in the present invention, the reality finally obtained smoothly goes out force curve can not delayed time delay, compared to the method utilizing first-order low-pass wave method to stabilize photovoltaic electric station fluctuation of power, the present invention has the effect of smoothly exerting oneself more optimized.

Accompanying drawing explanation

In order to be illustrated more clearly in the embodiment of the present invention or technical scheme of the prior art, below the accompanying drawing used required in embodiment being described is briefly described, apparently, accompanying drawing in the following describes is only some embodiments of the present invention, for those of ordinary skill in the art, under the prerequisite not paying creative work, it is also possible to obtain other accompanying drawing according to these accompanying drawings.

Fig. 1 is the system schematic of the wind energy turbine set that background of invention provides;

Fig. 2 be background of invention provide utilize first-order low-pass ripple algorithm carry out wind-powered electricity generation fluctuation stabilize before and after powertrace contrast schematic diagram;

Fig. 3 smoothly exerts oneself method flow schematic diagram in photovoltaic electric station provided by the invention;

Fig. 4 is that plan provided by the invention is exerted oneself the curve synoptic diagram of interval predicting power of photovoltaic plant value composition;

Fig. 5 is that the curve of interval predicting power of photovoltaic plant value composition is exerted oneself in plan provided by the invention and desired smooth is exerted oneself curve comparison schematic diagram;

The reality that Fig. 6 is corresponding diagram 5 provided by the invention smoothly goes out force curve.

Embodiment

Below in conjunction with the accompanying drawing in the embodiment of the present invention, the technical scheme in the embodiment of the present invention is clearly and completely described, it is clear that described embodiment is only the present invention's part embodiment, instead of whole embodiments. Based on the embodiment in the present invention, those of ordinary skill in the art, not making other embodiments all obtained under creative work prerequisite, belong to the scope of protection of the invention.

The present invention provides a kind of photovoltaic electric station smoothly to exert oneself method, and as shown in Figure 3, the method comprises:

Step S1, obtains the data acquisition of predicting power of photovoltaic plant value composition.

Concrete, this step can obtain power prediction value from the power prediction system SCADA at photovoltaic electric station, and the photovoltaic electric namely obtaining SCADA prediction stands in plan and exerts oneself the output power value of interval (following for some time). It is illustrated in figure 4 the curve that certain photovoltaic electric stands in the predicting power of photovoltaic plant value composition that plan is exerted oneself interval.

Step S2, utilizes polynomial fitting algorithm that data acquisition is carried out matching, obtains desired smooth and exerts oneself formula.

Concrete, this step utilizes polynomial fitting algorithm interval predicting power of photovoltaic plant Value Data of plan being exerted oneself to carry out matching, due to be not as existing utilize first-order low-pass ripple algorithm adopt adjacent filtering output value to calculate current filtering output value, therefore the desired smooth that the present invention obtains goes out force curve (i.e. desired smooth exert oneself curve corresponding to formula) and there will be no time delay phenomenon, and smooth effect is optimized more.

Step S3, goes out force value according to desired smooth formulae discovery desired smooth of exerting oneself.

Step S4, goes out force value and the magnitude relationship of predicting power of photovoltaic plant value, it is determined that reality smoothly goes out force value according to desired smooth.

Concrete, due to when light inadequate resource, photovoltaic electric station cannot increase exerts oneself, and that is, the power that photovoltaic DC-to-AC converter exports only has downward tunable characteristic, therefore, being greater than that photovoltaic electric station is actual exerts oneself when ideal is exerted oneself, namely desired smooth goes out force value when being greater than predicting power of photovoltaic plant value, and actual taking photovoltaic electric station is exerted oneself as control objectives generating, also it is exactly that reality smoothly goes out force value and equals predicting power of photovoltaic plant value, discharges peak power to electrical network; And when desirable exerting oneself is less than or equals the available of photovoltaic electric station and exert oneself, namely desired smooth goes out force value when being less than or equal predicting power of photovoltaic plant value, exert oneself as control objectives generating taking ideal, also it is exactly that reality smoothly goes out force value and equals desired smooth and go out force value.

Below the detailed process of above each step is described:

Step S1 is specially the data acquisition P obtaining predicting power of photovoltaic plant value composition:

P={ (p_i,t_i) | i=1,2..., m};

Wherein, P is data acquisition, p_iFor predicting power of photovoltaic plant value, m is the sample number of data acquisition and m is natural number, and i is sample sequence number, t_iFor p_iThe corresponding time.

Step S2 specifically comprises:

Step S21, according to predicting power of photovoltaic plant value p in data acquisition P_iFluctuation tendency, it is determined that desired smooth is exerted oneself the rank number n of formula, and wherein n is natural number.

Preferably, step S21 can specifically comprise:

According to predicting power of photovoltaic plant value p in data acquisition P_iFluctuation tendency, it is determined that desired smooth is exerted oneself curve waveform;

Exert oneself curve waveform according to described desired smooth, it is determined that described desired smooth is exerted oneself the rank number n of formula.

Such as, when described desired smooth exert oneself curve waveform be straight line time, it is determined that described desired smooth is exerted oneself the rank number n=1 of formula; When described desired smooth exert oneself curve waveform be para-curve time, it is determined that described desired smooth is exerted oneself the rank number n=2 of formula.

Step S22, matching has the polynomial expression of rank number n:

a_nt_i ⁿ+a_n-1t_i ^n-1+…+a₁t_i+a₀;

Wherein, a₀～a_nFor multinomial coefficient.

Step S23, evaluator a_nt_i ⁿ+a_n-1t_i ^n-1+…+a₁t_i+a₀With predicting power of photovoltaic plant value p_iSquared difference and Err:

$\mathrm{Err}=\underset{i=0}{\overset{m}{\mathrm{\Σ}}}{(a_n{t_i}^n+a_{n-1}{t_i}^{n-1}+...+a_1{t}_i+a_0-p_i)}^2.$

Step S24, when utilizing method of least squares calculated difference sum of squares Err for minimum value, multinomial coefficient a₀～a_nCorresponding concrete value α₀～α_n。

This step can adopt following account form:

Respectively to multinomial coefficient a₀～a_nSeek partial derivative, obtain following system of equations:

$\left\{\begin{array}{c}m{a}_0+\left(\underset{i=1}{\overset{m}{\mathrm{\Σ}}}{t}_i\right)a_1+...+\left(\underset{i=1}{\overset{m}{\mathrm{\Σ}}}{t_i}^n\right)a_n=\underset{i=1}{\overset{m}{\mathrm{\Σ}}}{p}_i\\ \left(\underset{i=1}{\overset{m}{\mathrm{\Σ}}}{t}_i\right)a_0+\left(\underset{i=1}{\overset{m}{\mathrm{\Σ}}}{t_i}^2\right)a_1+...+\left(\underset{i=1}{\overset{m}{\mathrm{\Σ}}}{t_i}^{n+1}\right)a_n=\underset{i=1}{\overset{m}{\mathrm{\Σ}}}{t}_i{p}_i\\ ...\\ ...\\ ...\\ \left(\underset{i=1}{\overset{m}{\mathrm{\Σ}}}{t_i}^n\right)a_0+\left(\underset{i=1}{\overset{m}{\mathrm{\Σ}}}{t_i}^{n+1}\right)a_1+...+\left(\underset{i=1}{\overset{m}{\mathrm{\Σ}}}{t_i}^{2n}\right)a_n=\underset{i=1}{\overset{m}{\mathrm{\Σ}}}{t_i}^n{p}_i\end{array}\right.$

Solve above system of equations, obtain multinomial coefficient a₀～a_nCorresponding concrete value α₀～α_n。

Step S25, utilizes and is specifically worth α₀～α_nBuild desired smooth to exert oneself formula X (t):

X (t)=α_ntⁿ+α_n-1t^n-1+…+α₁t+α₀;

Wherein, t is the time.

Step S3 is specially:

Calculate and work as t=t_iTime, desired smooth is exerted oneself the value X (t of formula X (t)_i):

X(t_i)=α_nt_i ⁿ+α_n-1t_i ^n-1+…+α₁t_i+α₀

Wherein, X (t_i) go out force value for desired smooth.

Step S4 specifically comprises:

When desired smooth goes out force value X (t_i) be less than or equal predicting power of photovoltaic plant value p_iTime, it is determined that reality smoothly goes out force value X (t_i) ' equal desired smooth goes out force value X (t_i), that is, as X (t_i)≤p_iTime, X (t_i) '=X (t_i)。

When desired smooth goes out force value X (t_i) it is greater than predicting power of photovoltaic plant value p_iTime, it is determined that reality smoothly goes out force value X (t_i) ' equal predicting power of photovoltaic plant value p_i, that is, as X (t_i)>p_iTime, X (t_i) '=p_i。

Fig. 5 respectively illustrates plan and exerts oneself the curve (shown in dotted line) of interval predicting power of photovoltaic plant value composition and the desired smooth that utilizes method provided by the invention to calculate goes out force curve (shown in solid line), by contrasting, desired smooth goes out force curve and reduces fluctuation of power, and there is not time delay phenomenon. Go out the magnitude relationship of force value according to moment predicting power of photovoltaic plant value each in Fig. 5 and desired smooth, obtain the reality shown in Fig. 6 and smoothly go out force curve.

Above specific embodiment; the object of the present invention, technical scheme and useful effect have been further described; it is it should be understood that; these are only specific embodiments of the invention; the protection domain being not intended to limit the present invention; within the spirit and principles in the present invention all, any amendment of making, equivalent replacement, improvement etc., all should be included within protection scope of the present invention.

Claims (5)

1. smoothly exert oneself method in a photovoltaic electric station, it is characterised in that, comprising:

Steps A, obtains the data acquisition of predicting power of photovoltaic plant value composition;

Step B, utilizes polynomial fitting algorithm that described data acquisition is carried out matching, obtains desired smooth and exerts oneself formula;

Step C, goes out force value according to described desired smooth formulae discovery desired smooth of exerting oneself;

Step D, goes out force value and the magnitude relationship of described predicting power of photovoltaic plant value, it is determined that reality smoothly goes out force value according to described desired smooth;

Described steps A is specially the data acquisition P obtaining predicting power of photovoltaic plant value composition:

P={ (p_i,t_i) | i=1,2..., m};

Wherein, P is data acquisition, p_iFor predicting power of photovoltaic plant value, m is the sample number of described data acquisition and m is natural number, and i is sample sequence number, t_iFor p_iThe corresponding time;

Described step B specifically comprises:

Step B1, according to predicting power of photovoltaic plant value p in described data acquisition P_iFluctuation tendency, it is determined that described desired smooth is exerted oneself the rank number n of formula, and wherein n is natural number;

Step B2, matching has the polynomial expression of described rank number n:

a_nt_i ⁿ+a_n-1t_i ^n-1+L+a₁t_i+a₀;

Wherein, a₀～a_nFor multinomial coefficient;

Step B3, calculates described polynomial expression a_nt_i ⁿ+a_n-1t_i ^n-1+L+a₁t_i+a₀With described predicting power of photovoltaic plant value p_iSquared difference and Err:

$\mathrm{Err}=\underset{i=1}{\overset{m}{\mathrm{\Σ}}}{(a_n{t_i}^n+a_{n-1}{t_i}^{n-1}+\·\·\·+a_1{t}_1+a_0-p_i)}^2;$

Step B4, when utilizing method of least squares to calculate described squared difference and Err for minimum value, multinomial coefficient a₀～a_nCorresponding concrete value α₀～α_n;

Step B5, utilizes described concrete value α₀～α_nBuild desired smooth to exert oneself formula X (t):

X (t)=α_ntⁿ+α_n-1t^n-1+L+α₁t+α₀;

Wherein, t is the time;

Described step C is specially:

Calculate and work as t=t_iTime, described desired smooth is exerted oneself the value X (t of formula X (t)_i):

X(t_i)=α_nt_i ⁿ+α_n-1t_i ^n-1+L+α₁t_i+α₀

Wherein, X (t_i) go out force value for desired smooth;

Described step D specifically comprises:

When described desired smooth go out force value be less than or equal described predicting power of photovoltaic plant value time, it is determined that reality smoothly goes out force value and equals described desired smooth and go out force value;

When described desired smooth go out force value be greater than described predicting power of photovoltaic plant value time, it is determined that reality smoothly goes out force value and equals described predicting power of photovoltaic plant value.

2. method according to claim 1, it is characterised in that, described step B1 specifically comprises:

According to predicting power of photovoltaic plant value p in data acquisition P_iFluctuation tendency, it is determined that desired smooth is exerted oneself curve waveform;

Exert oneself curve waveform according to described desired smooth, it is determined that described desired smooth is exerted oneself the rank number n of formula.

3. method according to claim 2, it is characterised in that, when described desired smooth exert oneself curve waveform be straight line time, it is determined that described desired smooth is exerted oneself the rank number n=1 of formula.

4. method according to claim 2, it is characterised in that, when described desired smooth exert oneself curve waveform be para-curve time, it is determined that described desired smooth is exerted oneself the rank number n=2 of formula.

5. method according to claim 1, it is characterised in that, described step B4 specifically comprises:

Respectively to multinomial coefficient a₀～a_nSeek partial derivative, obtain following system of equations:

$\left\{\begin{array}{c}{\mathrm{ma}}_0+\left(\underset{i=1}{\overset{m}{\mathrm{\Σ}}}{t}_i\right)a_1+\·\·\·+\left(\underset{i=1}{\overset{m}{\mathrm{\Σ}}}{t_i}^n\right)a_n=\underset{i=1}{\overset{m}{\mathrm{\Σ}}}{p}_i\\ \left(\underset{i=1}{\overset{m}{\mathrm{\Σ}}}{t}_i\right)a_0+\left(\underset{i=1}{\overset{m}{\mathrm{\Σ}}}{t_i}^2\right)a_1+\·\·\·+\left(\underset{i=1}{\overset{m}{\mathrm{\Σ}}}{t_i}^{n+1}\right)a_n=\underset{i=1}{\overset{m}{\mathrm{\Σ}}}{t}_i{p}_i\\ \·\·\·\\ \·\·\·\\ \·\·\·\\ \left(\underset{i=1}{\overset{m}{\mathrm{\Σ}}}{t_i}^n\right)a_0+\left(\underset{i=1}{\overset{m}{\mathrm{\Σ}}}{t_i}^{n+1}\right)a_1+\·\·\·+\left(\underset{i=1}{\overset{m}{\mathrm{\Σ}}}{t_i}^{2n}\right)a_n=\underset{i=1}{\overset{m}{\mathrm{\Σ}}}{t_i}^n{p}_i\end{array}\right.$

Solve above system of equations, obtain multinomial coefficient a₀～a_nCorresponding concrete value α₀～α_n。