CN102684222B - Method for smoothly controlling wind power generation power based on energy storage technology - Google Patents

Abstract

The invention discloses a method for smoothly controlling wind power generation power based on an energy storage technology. The method comprises the following steps of: firstly analyzing the output power of a wind power station through fast Fourier transform (FFT) so as to obtain an amplitude-frequency characteristic curve of the output power of the wind power station; extracting an expected grid-tied power and a high-frequency signal power from the output power of the wind power station through wavelet transform (WT); absorbing the high-frequency signal power by virtue of a storage battery energy storage system and a supercapacitor; when the output power of the wind power station is less than the expected grid-tied power signal, controlling the storage battery energy storage system and the supercapacitor to discharge by using a model algorithm; when the output power of the wind power station is greater than the expected grid-tied power, controlling the storage battery energy storage system and the supercapacitor to charge by using a model algorithm; and giving the optimal volume configuration of the storage battery energy storage system by using WT. The method disclosed by the invention can be used for improving the safety and economical efficiency of grid-tied operations.

Description

A kind of wind power generation power smooth control method based on energy storage technology

Technical field

The invention belongs to wind storing cogeneration technical field, relate in particular to a kind of wind power generation power smooth control method based on energy storage technology.

Background technology

Along with the progress of world energy sources exploitation, regenerative resource more and more obtains people's favor, and wherein the development of wind power generation obtains more people's concern.Because wind speed has, can not expect and the feature such as random fluctuation, the fluctuation meeting of wind-powered electricity generation unit power output produces larger impact to the quality of power supply, as voltage deviation, voltage fluctuation and flickering, harmonic wave etc.Wind-electricity integration power fluctuation is large, and this brings negative effect to the stability of operation of power networks and economy.Therefore, the inhibition problem of research output power fluctuation of wind farm has important practical significance.

The fluctuation that at present adopts energy storage device to stabilize Power Output for Wind Power Field not only can regulate reactive power, also can regulating power, can effectively suppress output pulsation, and obtain good smooth effect.From pumped storage and compressed-air energy storage in Power Output for Wind Power Field should be used for aspect level and smooth, although these two kinds of technology are comparatively ripe at present, but when wind energy turbine set capacity reaches megawatt hours up to ten thousand, these two kinds of technology all require very high to natural environment, construction bureau is sex-limited large, and construction investment in early stage is quite huge, and dynamic adjustments response speed is slow, is not suitable for large-scale wind power field and uses.Utilize flywheel energy storage system can improve the quality of power supply and the stability of grid connected wind power field, flywheel energy storage service life cycle is long, safeguard simple, can continuous operation, there is good development prospect, but at aspects such as rotor strength design, low-watt consumption magnetic bearing, security protections, there are technological difficulties at present, be badly in need of breaking through.Superconducting magnetic energy storage is applied to the output of smooth wind power field system, and superconducting energy storage technology itself has a series of advantage, but price comparison is expensive, and application is not at present very extensive.Ultracapacitor is a kind of outstanding energy-storage travelling wave tube between traditional physical capacitor and battery, have that probability density is high, charge and discharge circulation life is long, the charging interval is short, storage life is long, high reliability, but it is still not high enough that its technological difficulties are voltage endurance capability, even the withstand voltage level of ceramic ultracapacitor is the highest at present, also can only bear 1kV left and right, and cost is high.In recent years, the still electrochemical energy storage technology being most widely used, particularly in small distributed electricity generation system, lead acid accumulator energy density is moderate, low price, constructions cost are low, technology maturation, security performance are relatively reliable, but its cycle life is shorter, can not deep discharge, after the circulation several years waste battery harmless treatment and to the drawbacks limit such as harm of environment its use.Sodium-sulphur battery energy density is high, has extended cycle life, and the successful case of electrochemical energy storage is all to use this technology in the world.Flow battery is mainly by the electrolyte solution and battery module and the corresponding composition of the control system that are contained in different storage tanks, there is the advantages such as security reliability is high, electrolyte can recycle, the theoretical life-span is long, it is low that but it also has energy density, volume is large, floor space is huge, too high to ambient temperature requirement, price, the shortcomings such as systematic comparison complexity.Flow battery has a variety of, most widely used, the most ripe flow battery is vanadium redox flow battery at present, the vanadium cell speed of response is fast, can under peak load, move continuously, and can non-hazardously discharge completely, but its shortcoming is can be to environment, and not commercialization, if can ripely develop, will be well suited for China's large-scale wind power energy storage, there are wide market prospects.

From studying above, energy storage technology is varied, and respectively has its own feature and unique application background.Cooperation is with corresponding control strategy, and energy-storage system just can effectively be realized the level and smooth of output power fluctuation of wind farm.At present conventional digital filter and the PI method power of realizing wind storage system that combines is smoothly controlled, while finding to adopt digital filter to stabilize the fluctuation of Power Output for Wind Power Field by emulation experiment, simulation curve significantly lags behind actual output power curve (as shown in Figure 5), in the situation that power changes greatly, not only make energy-storage system energy adjusting amplitude large, adjustment cycle is also longer.Meanwhile, the digital filter time constant in the method and power smooth effect and stored energy capacitance dispose direct relation, time constant is chosen too large, needs the stored energy capacitance that configuration is larger, has increased energy-storage system cost; Time constant is chosen too little, need the stored energy capacitance of configuration less, but wind power fluctuation is larger.Therefore, grid-connected on a large scale along with wind-powered electricity generation, traditional control method can not meet its requirement in some aspects, and for the better effectiveness of performance energy-storage system, the how tactful energy-storage system of mixing will provide a new thinking for us.

Power based on hybrid energy-storing technology is smoothly controlled, fully analyzing on the basis of Power Output for Wind Power Field amplitude-frequency characteristic, by wavelet transformation, Power Output for Wind Power Field signal is carried out to multiple dimensioned decomposition, obtain having more low frequency signal and the high-frequency signal of cyclophysis, and the hybrid energy-storing mode of selection and its adaptation is carried out the level and smooth of Power Output for Wind Power Field, set up the wind power smoothing model of mixed energy storage system, adopt model algorithm to control the control that discharges and recharges that realizes energy-storage battery, reach the object that power is level and smooth, under the prerequisite that meets wind-electricity integration standard, obtain optimum stored energy capacitance configuration, reduce energy storage device scale, reduce energy storage cost, it is a kind of novelty, the reliable but level and smooth and stored energy capacitance Optimal Configuration Method of power accurately.

Summary of the invention

For the feature of single type energy-storage system of mentioning in above-mentioned background technology and the deficiency that Power Output for Wind Power Field smoothing method exists at present, the present invention proposes a kind of wind power generation power smooth control method based on energy storage technology.

Technical scheme of the present invention is that a kind of wind power generation power smooth control method based on energy storage technology, is characterized in that the method comprises the following steps:

Step 1: change Power Output for Wind Power Field is analyzed by fast Fourier, obtain the amplitude-versus-frequency curve of Power Output for Wind Power Field;

Step 2: on the basis of step 1, extract the grid-connected power of expectation and high-frequency signal power from Power Output for Wind Power Field;

Step 3: absorb described high-frequency signal power by energy-storage system of accumulator and ultracapacitor;

Step 4: when the power output of wind energy turbine set is less than the grid-connected power signal of described expectation, control energy-storage system of accumulator and ultracapacitor electric discharge; When the power output of wind energy turbine set is greater than the grid-connected power of described expectation, control energy-storage system of accumulator and ultracapacitor charging.

After described step 4, also comprise the optimum capacity configuration of calculating accumulator energy-storage system.

The extraction formula of the grid-connected power of described expectation is:

Wherein:

for representing the grid-connected power of expectation;

C _{j, n}for yardstick expansion coefficient;

for SPACE V _jorthonormal basis.

The extraction formula of described high-frequency signal power is:

$G_d^j\left(t\right)=\underset{n}{\mathrm{\Σ}}{d}_{j,n}{\mathrm{\φ}}_{j,n}\left(t\right)$

Wherein:

for representing high-frequency signal power;

D _{j, n}for Wavelet Expansions coefficient;

φ _{j, n}(t) be space W _jorthonormal basis.

The computing formula of the optimum capacity of described energy-storage system of accumulator is:

$E=\max {\∫}_0^t{y}_0\left(t\right)\mathrm{dt}$

Wherein:

E is the optimum capacity of energy-storage system of accumulator;

Y ₀(t) expect that in advance power x's (t) is poor for time domain wind power output power g (t) and time domain;

for the energy storage energy that energy-storage system absorbs or emits within 0 to t time.

The invention has the beneficial effects as follows and propose a kind of Power Output for Wind Power Field smooth control method based on hybrid energy-storing technology, the method can make full use of the feature of various energy storage devices, on the basis based on wavelet decomposition, obtains and is suitable for the signal characteristic that different energy storage devices have.Adopt model algorithm to control and fast energy-storage system of accumulator is discharged and recharged to control, make the grid-connected performance number of the more approaching expectation of output of mixed energy storage system, compare single type energy-storage system, hybrid energy-storing technology is low-yield to high frequency, the high-octane power fluctuation of low frequency all can be realized good smooth effect.What while energy-storage system under wavelet transformation mode was level and smooth is the energy of HFS, because its average is zero, can obtain optimum stored energy capacitance configuration, greatly reduces the operating cost of energy-storage system.The level and smooth control strategy of Power Output for Wind Power Field that utilizes the inventive method to obtain, can, for wind storing cogeneration system, improve the fail safe and the economy that are incorporated into the power networks.The present invention compares with existing method, is a kind of new and power smooth control method accurately and reliably.

Accompanying drawing explanation

Fig. 1 is Power Output for Wind Power Field amplitude-versus-frequency curve;

Fig. 2 is the multiple dimensioned decomposition curve of Power Output for Wind Power Field signal based on wavelet transformation; Fig. 2 a is original power signal and low frequency A thereof ₇and high frequency D (t) ₇(t) curve; Fig. 2 b is high frequency D ₆(t), D ₅(t), D ₄(t) curve; Fig. 2 c is high frequency D ₃(t), D ₂(t), D ₁(t) curve;

Fig. 3 is that the Power Output for Wind Power Field based on hybrid energy-storing technology is smoothly controlled model;

Fig. 4 is that the Power Output for Wind Power Field of hybrid energy-storing technology is smoothly controlled Simulink emulation platform and model algorithm control flow chart; Fig. 4 a is that the Power Output for Wind Power Field of hybrid energy-storing technology is smoothly controlled Simulink emulation platform; Fig. 4 b is the flow chart that model algorithm is controlled;

Fig. 5 is that the Power Output for Wind Power Field based on wavelet transformation and model algorithm control is smoothly controlled simulation curve;

Fig. 6 is the battery energy storage capacity configuration based on wavelet transformation.

Embodiment

Below in conjunction with accompanying drawing, preferred embodiment is elaborated.Should be emphasized that, following explanation is only exemplary, rather than in order to limit the scope of the invention and to apply.

The present invention carries out multiple dimensioned decomposition by the multiresolution analysis method in wavelet theory to Power Output for Wind Power Field, combining super capacitor device and energy-storage system of accumulator Energy distribution feature, and the mixed energy storage system power having built based on wavelet transformation is smoothly controlled model.In this model, the low frequency part of Power Output for Wind Power Field after wavelet decomposition is as the grid-connected performance number of expectation, and HFS is absorbed by ultracapacitor and energy-storage system of accumulator respectively.Wherein, according to battery response speed choose reasonable time high frequency power signal, by the capacity energy-storage system of accumulator slower compared with high response speed, undertaken smoothly, incomplete smooth and remaining HFS ultracapacitor little by capacity, fast response time are auxiliary level and smooth.Then the grid-connected power signal after the charge and discharge that adopt model algorithm control to realize rapidly energy-storage system of accumulator control to guarantee is smoothly followed the tracks of desired wind-electricity integration performance number better, realizes the level and smooth object of controlling of Power Output for Wind Power Field.Simultaneously, meeting under the prerequisite of wind-electricity integration standard, the actual power of wind energy turbine set output and the difference of the grid-connected power of resulting expectation after wavelet filtering are quadratured, calculate the maximum stored energy capacitance of the required configuration of energy-storage system of accumulator under different time, adopt the stored energy capacitance collocation method providing in the present invention can greatly reduce the operation and maintenance cost of energy-storage system.

The inventive method comprises the following steps:

Step 1: change Power Output for Wind Power Field is analyzed by fast Fourier, obtain the amplitude-versus-frequency curve of Power Output for Wind Power Field;

Step 2: on the basis of step 1, extract the grid-connected power of expectation and high-frequency signal power by wavelet transformation from Power Output for Wind Power Field;

Step 3: by energy-storage system of accumulator and ultracapacitor absorbing high-frequency signal power;

Step 4: when the power output of wind energy turbine set is less than the grid-connected power signal of expectation, control energy-storage system of accumulator and ultracapacitor electric discharge with model algorithm; When the power output of wind energy turbine set is greater than the grid-connected power of expectation, with model algorithm, control energy-storage system of accumulator and ultracapacitor charging;

Afterwards, with wavelet transformation, provide the optimum capacity configuration of energy-storage system of accumulator.

The level and smooth degree of Power Output for Wind Power Field is weighed by the fluctuation ratio of Power Output for Wind Power Field, in < < wind energy turbine set access electric power network technique regulation > >, require the wind energy turbine set that total installation of generating capacity is 100MW, the maximum power variation rate allowing in its 1 minute should be less than 20%, and its computing formula is:

$\mathrm{\&gamma;}=\frac{x\left(t\right)-x(t-\mathrm{\&Delta;t})}{P_m}\&times;100\%$

Wherein:

γ is fluctuation ratio;

X (t) is the t grid-connected performance number of expectation constantly;

X (t-Δ t) is the t-Δ t grid-connected performance number of expectation constantly;

P _minstalled capacity for wind energy turbine set.

The present invention is divided into following components:

(1) Analysis of Magnitude-Frequency Characteristic of Power Output for Wind Power Field: obtain certain 99MW wind energy turbine set in April, 2010 and annual real output data, directly call FFT function in Matlab tool box and these data are carried out to fast Fourier transform obtain amplitude-versus-frequency curve, show that the energy of Power Output for Wind Power Field mainly concentrates on low frequency part (0～10 ^-4hz), the lower conclusion of its HFS energy.

(2) the multiple dimensioned decomposition of Power Output for Wind Power Field signal: known by the Analysis of Magnitude-Frequency Characteristic in (1), the energy of Power Output for Wind Power Field mainly concentrates on low frequency part, therefore using low frequency power signal as wind-powered electricity generation, expect grid-connected performance number, high frequency power signal is undertaken smoothly by energy-storage system.Utilize the multiresolution analysis method in wavelet transformation Power Output for Wind Power Field signal decomposition can be become to high-frequency signal and low frequency signal, adopt the DB9 small echo in wavelet transformation to carry out 7 layers of decomposition to wind power output power signal here, obtain its low frequency part A ₇and D (t) _i(t) (i=1,2 ... 7) power curve.Low frequency part A ₇(t) be the leading part of original power curve, HFS D ₇(t) energy is less, and near fluctuation up and down null value, utilizes the energy of level and smooth this HFS of this feature, can reach the level and smooth object of power.

(3) Power Output for Wind Power Field based on hybrid energy-storing technology is smoothly controlled the foundation of model: the advantage of comprehensive utilization energy-storage system of accumulator and ultracapacitor, the high frequency power signal D after wavelet decomposition in (2) ₇(t) by the capacity energy-storage system of accumulator slower compared with high response speed, absorbed, due to the inertia of battery and the impact of response speed, the actual output H of battery ₀(s) be not equal to desired value Y ₀(s), incomplete absorption portion and other HFS D ₁(t)～D ₆(t) the auxiliary absorption of ultracapacitor little by capacity, fast response time, finally makes the output C (s) of mixed energy storage system more approach desired value X (s), and the power of having set up based on hybrid energy-storing technology is smoothly controlled model.

(4) adopt model algorithm to control the control that discharges and recharges that realizes mixed energy storage system: model algorithm is controlled and claimed again model prediction to inspire control, mainly comprises several parts such as setting up internal model, feedback compensation, rolling optimization.Its control thought is as follows:

First, utilize the input-output characteristic that energy-storage system of accumulator is known, set up its nonparametric model

wherein y ' (k) represents k predicted value constantly, before constantly N of this value and k input and object impulse response coefficient relevant; The length of N indicated object impulse response, thinks that the input in top n that object is only subject to constantly affects; Impulse response H (l)=10 ³* [9.99.89.7 ... ], l=1,2 ..., N.

Secondly, utilize the power stage value y ' in this model prediction energy-storage system of accumulator future (k), and according to predicted value y ' (k) with the deviation e (k) of actual value y (k), use feedback correcting coefficient β to revise predicted value, wherein y (k) represents k value constantly, β=0.1;

Then, ask for and calculate predicted value Y _p(k) with reference locus Y _r(k) deviation, wherein, Y _pand Y (k) _r(k) represent the value of a time period, α is constant, gets 0.1 here;

Finally, use quadratic form performance objective function $\min J={\left|\right|Y_p\left(k\right)-Y_r\left(k\right)\left|\right|}_{\mathrm{<figure-callout\; id="2"\; label="Q"\; filenames="HDA00001637165000011.png,HDA00001637165000032.png"\; state="\{\{state\}\}">Q</figure-callout>}}^2+{\left|\right|U\left(k\right)\left|\right|}_{\mathrm{<figure-callout\; id="2"\; label="R"\; filenames="HDA00001637165000011.png,HDA00001637165000032.png"\; state="\{\{state\}\}">R</figure-callout>}}^2$ Calculate current optimal control amount U (k), wait until the next sampling period, repeat this process.

In step (1), adopt fast fourier transform to obtain the amplitude-versus-frequency curve of Power Output for Wind Power Field, wind power energy mainly concentrates on its low frequency part, and HFS energy is lower.This and wind speed characteristics match, and the wind speed amplitude that high frequency changes is very little, and low frequency variations wind speed amplitude is larger.Utilize the energy of corresponding its HFS of filtering algorithm filtering, retain the energy of its low frequency part, meet grid-connected power level and smooth time, take into account the impact on energy-storage system performance.Therefore, adopt the multiresolution analysis method in wavelet transformation that Power Output for Wind Power Field signal decomposition is become to high-frequency signal and low frequency signal.If g (t) ∈ is L ²(R) be Power Output for Wind Power Field signal to be decomposed, by multiresolution Analysis Theory, obtained:

The extraction formula of expecting grid-connected power is:

Wherein:

for representing the grid-connected power of expectation;

C _{j, n}for yardstick expansion coefficient;

for SPACE V _jorthonormal basis.

The extraction formula of high-frequency signal power is:

$g_d^j\left(t\right)=\underset{n}{\mathrm{\&Sigma;}}{d}_{j,n}{\mathrm{\&phi;}}_{j,n}\left(t\right)$

Wherein:

for representing high-frequency signal power;

D _{j, n}for Wavelet Expansions coefficient;

φ _{j, n}(t) be space W _jorthonormal basis.

Power Output for Wind Power Field signal decomposes the low frequency signal obtaining by wavelet transformation, because its energy is high, change slowly, play a leading role, will it as expecting grid-connected performance number, and higher frequency signal energy is low, it is fast to change, it is undertaken smoothly by ultracapacitor and energy-storage system of accumulator respectively.Because storage battery has the features such as the high and power density of energy density, service life cycle are low, high-frequency signal is undertaken smoothly by energy-storage system of accumulator; Due to the feature such as ultracapacitor has power density, cycle life is high and energy density is low, the power signal not smoothed out by energy-storage system of accumulator and other high-frequency signal are sent to ultracapacitor and carry out quick filter processing, the Power Output for Wind Power Field of having set up based on hybrid energy-storing technology is smoothly controlled model.

When the actual active power G of wind energy turbine set (s) is greater than the grid-connected power X of expectation (s) that wavelet transformation provides, i.e. X (s)-G (s)=Y ₀(s) < 0, by model algorithm, controlled energy-storage system of accumulator and ultracapacitor are sent to charging instruction; When the actual active power G of wind energy turbine set (s) is less than the grid-connected power X of expectation (s) that wavelet transformation provides, i.e. X (s)-G (s)=Y ₀(s) > 0, by model algorithm, controlled energy-storage system of accumulator and ultracapacitor are sent to electric discharge instruction.Due to the inertia of battery and the impact of response speed, the actual output H of battery ₀(s) be not equal to desired value Y ₀(s), the actual inferior HFS that discharges and recharges instruction that completes of energy-storage system of accumulator, discharges and recharges the HFS of instruction and energy-storage system of accumulator and discharges and recharges deviation part and by ultracapacitor, be responsible for level and smooth.

Stored energy capacitance is the energy that the energy-storage system of accumulator in a period of time is level and smooth or discharge, and is the integration of energy storage power within this period.Definite method of stored energy capacitance has a variety of, and frequency-region signal is transformed into time domain, in the present invention to the actual power g (t) of wind energy turbine set output the difference y with the grid-connected power x of resulting expectation (t) after wavelet transform filtering ₀(t) quadrature, calculate maximum integral value under different time as optimum stored energy capacitance, that is:

$E=\max {\&Integral;}_0^t{y}_0\left(t\right)\mathrm{dt}$

Wherein:

E is the optimum capacity of energy-storage system of accumulator;

Y ₀(t) expect that in advance power x's (t) is poor for time domain wind power output power g (t) and time domain;

for the energy storage energy that energy-storage system absorbs or emits within 0 to t time.

Fig. 1 is Power Output for Wind Power Field amplitude-versus-frequency curve, and as shown in Figure 1, Power Output for Wind Power Field obtains its amplitude-versus-frequency curve through fast fourier transform.Power Output for Wind Power Field energy mainly concentrates on its low frequency part (0～10 ^-4and HFS energy is lower Hz).

Fig. 2 is the multiple dimensioned decomposition curve of Power Output for Wind Power Field signal based on wavelet transformation, and Fig. 2 a is original power signal and low frequency A thereof ₇and high frequency D (t) ₇(t) curve; Fig. 2 b is high frequency D ₆(t), D ₅(t), D ₄(t) curve; Fig. 2 c is high frequency D ₃(t), D ₂(t), D ₁(t) curve;

As shown in Figure 2, adopt DB9 wavelet function to carry out 7 layers of decomposition to Power Output for Wind Power Field signal, obtain its low frequency part (1/60/27=1.3*10 ^-4hz) power curve, the equation before decomposing and after decomposing closes and is: S=A ₇(t)+D ₇(t)+D ₆(t)+D ₅(t)+D ₄(t)+D ₃(t)+D ₂(t)+D ₁(t), low frequency A ₇(t) be the leading part of original power curve, as the desired grid-connected performance number X (s) of wind-powered electricity generation, and HFS D ₁(t)-D ₇(t) be the principal element that affects output power fluctuation of wind farm, adopt ultracapacitor and model algorithm control system to be responsible for level and smooth.

Fig. 3 is that the Power Output for Wind Power Field based on hybrid energy-storing technology is smoothly controlled model.The advantage of comprehensive utilization energy-storage system of accumulator and ultracapacitor, the inferior high frequency power signal D after wavelet decomposition ₇(t) by the capacity energy-storage system of accumulator slower compared with high response speed, absorbed, due to the inertia of storage battery and the impact of response speed, the actual output H of battery ₀(s) be not equal to desired value Y ₀(s), incomplete absorption portion and other HFS Y ₀(s)-H ₀(s)=Y ₁(s) the auxiliary absorption of ultracapacitor little by capacity, fast response time, H ₁(s) be the output valve of ultracapacitor.Finally make the output C (s) of mixed energy storage system more approach desired value X (s), i.e. C (s)=G (s)+H ₀(s)+H ₁(s)=G (s)+Y ₀(s)=X (s), the power of having set up based on hybrid energy-storing technology is smoothly controlled model.

Fig. 4 is that the Power Output for Wind Power Field of hybrid energy-storing technology is smoothly controlled Simulink emulation platform and model algorithm control flow chart.As shown in Fig. 4 a, the power of having built based on hybrid energy-storing technology is smoothly controlled Simulink emulation platform.Adopt model algorithm to control energy-storage system of accumulator is discharged and recharged to control, control flow as shown in Figure 4 b.

Fig. 5 is that the Power Output for Wind Power Field based on wavelet transformation and model algorithm control is smoothly controlled simulation curve.In Fig. 5, the Power Output for Wind Power Field having provided under two kinds of methods is smoothly controlled curve.Simulation curve (the inertial rate ripple time constant T=1800s wherein that comprises the Power Output for Wind Power Field smooth control method that the digital filter mentioned in background technology and PI combine, parameter P=5, the Ts=0.1s of PI) and control the level and smooth simulation curve of controlling of the power combining based on wavelet transformation and model algorithm.As can be seen from the figure, digital filter method exists obvious hysteresis, adopt PI control algolithm (P=5, Ts=0.1) controlling energy-storage battery carries out while discharging and recharging task, impact due to controlled algorithm and energy-storage battery characteristic, the level and smooth degree of networking power that actual networking power is compared expection declines, and curve of output exists certain fluctuation, and its maximum fluctuation rate is 2.07%; And adopt model algorithm to control to have taken into account robustness good and accuracy advantages of higher, rapidly the input desired value Y of tracing energy-storage battery ₀(s), obtain more level and smooth power output, do not have hysteresis, its maximum fluctuation rate is only 0.58%.

Fig. 6 is the battery energy storage capacity configuration based on wavelet transformation.Two kinds of stored energy capacitance collocation methods based on inertial rate ripple and wavelet transformation in Fig. 6, have been provided equally.As can be seen from the figure, because digital filter method has obvious hysteresis error, get back to the initial condition that energy-storage system of accumulator discharges and recharges slower, capacity configuration in 50 hours has reached 40.58MWh, and curve after wavelet transformation all fluctuates near the initial condition of energy-storage system of accumulator, its energy storage system capacity configuration is only 21.21MWh.Owing to there is bad value in original power data, there is larger volume change in energy-storage system between 30h-40h, because actual power bust is to due to zero, if reject the bad value at this place, and can be less by the stored energy capacitance of the required configuration of small wave converting method.Table 1 and table 2 have provided respectively different time and annual each month needed stored energy capacitance configuration result in April.More known by two kinds of methods, meeting under the prerequisite of wind-electricity integration standard, the method adopting in the present invention has reduced the stored energy capacitance of required configuration greatly, has reduced the operating cost of energy-storage system.

Table 1 is the stored energy capacitance configuration of certain wind energy turbine set month different time;

Certain month battery capacity configuration (Level=7, ' db9 ' of table 1)

Table 2 is the capacity configurations of annual each month of certain wind energy turbine set;

Table 2 each month certain year battery capacity configuration and fluctuation ratio (Level=7, ' db9 ')

The level and smooth control of the power of the present invention in wind storing cogeneration system and stored energy capacitance have clear superiority in distributing rationally.

The above; be only the present invention's embodiment preferably, but protection scope of the present invention is not limited to this, is anyly familiar with in technical scope that those skilled in the art disclose in the present invention; the variation that can expect easily or replacement, within all should being encompassed in protection scope of the present invention.Therefore, protection scope of the present invention should be as the criterion with the protection range of claim.

Claims (5)

1. the wind power generation power smooth control method based on energy storage technology, is characterized in that the method comprises the following steps:

Step 1: change Power Output for Wind Power Field is analyzed by fast Fourier, obtain the amplitude-versus-frequency curve of Power Output for Wind Power Field;

Step 2: on the basis of step 1, adopt small wave converting method to carry out multiple dimensioned decomposition to Power Output for Wind Power Field signal, extract the grid-connected power of expectation and high-frequency signal power from Power Output for Wind Power Field;

Step 3: absorb described high-frequency signal power by energy-storage system of accumulator and ultracapacitor, thereby the power of setting up based on hybrid energy-storing technology is smoothly controlled model, makes the output C (s) of mixed energy storage system more approach desired value X (s); Detailed process is:

High frequency power signal after wavelet decomposition in step 2 is absorbed by energy-storage system of accumulator, due to the inertia of storage battery and the impact of response speed, and the actual output H of storage battery ₀(s) be not equal to desired value Y ₀(s), incomplete absorption portion and other HFS Y ₀(s)-H ₀(s)=Y ₁(s) by ultracapacitor is auxiliary, absorb;

Step 4: adopt model algorithm to obtain current optimal control amount U (k), using optimal control amount U (k) as discharging and recharging instruction, realize the control that discharges and recharges of mixed energy storage system; When the power output of wind energy turbine set is less than the grid-connected power signal of described expectation, control energy-storage system of accumulator and ultracapacitor electric discharge; When the power output of wind energy turbine set is greater than the grid-connected power of described expectation, control energy-storage system of accumulator and ultracapacitor charging; The detailed process of model control algolithm is:

First, utilize the input-output characteristic that energy-storage system of accumulator is known, set up its nonparametric model y'(k wherein) represent k predicted value constantly, before constantly N of this value and k input and object impulse response coefficient relevant; The length of N indicated object impulse response, the input of the top n that object is only subject in the moment affects; Impulse response H (l)=10 ³* [9.9 9.8 9.7 ... ], l=1,2 ..., N;

Secondly, utilize this model prediction energy-storage system of accumulator power stage value y'(k in future), and according to predicted value y'(k) with the deviation e (k) of actual value y (k), use feedback correcting coefficient β to revise predicted value, wherein y (k) represents k value constantly;

Then, ask for and calculate predicted value Y _p(k) with reference locus Y _r(k) deviation, wherein, Y _pand Y (k) _r(k) represent the value of a time period;

Finally, use quadratic form performance objective function

calculate current optimal control amount U (k), wait until the next sampling period, repeat this process; The charge and discharge instruction that wherein U (k) is mixed energy storage system.

2. a kind of wind power generation power smooth control method based on energy storage technology according to claim 1, is characterized in that also comprising the optimum capacity configuration of calculating accumulator energy-storage system after described step 4.

3. a kind of wind power generation power smooth control method based on energy storage technology according to claim 1, is characterized in that the extraction formula that described employing small wave converting method extracts the grid-connected power of expectation from Power Output for Wind Power Field is:

Wherein:

(t) for representing the grid-connected power of expectation;

C _j,nfor yardstick expansion coefficient;

(t) be SPACE V _jorthonormal basis.

4. a kind of wind power generation power smooth control method based on energy storage technology according to claim 1, is characterized in that the extraction formula that described employing small wave converting method extracts high-frequency signal power from Power Output for Wind Power Field is:

$G_d^j\left(t\right)=\underset{n}{\mathrm{\&Sigma;}}{d}_{j,n}{\mathrm{\&phi;}}_{j,n}\left(t\right)$

Wherein:

(t) for representing high-frequency signal power;

D _j,nfor Wavelet Expansions coefficient;

φ _j,n(t) be space W _jorthonormal basis.

5. a kind of wind power generation power smooth control method based on energy storage technology according to claim 2, is characterized in that the computing formula of the optimum capacity of described energy-storage system of accumulator is:

$E=\max {\&Integral;}_0^t{y}_{0\left(t\right)\mathrm{dt}}$

Wherein:

E is the optimum capacity of energy-storage system;

Y ₀(t) expect that in advance power x's (t) is poor for time domain wind power output power g (t) and time domain;

y ₀(t) dt is the energy storage energy that energy-storage system absorbs or emits within 0 to t time.

Images