CN105305413A - Wind and photovoltaic complementation generation system optimization configuration method - Google Patents

Abstract

The invention discloses a wind and photovoltaic complementation generation system optimization configuration method. The method comprises steps that, correlation between wind generation and photovoltaic generation is calculated to acquire wind and photovoltaic generation output joint probability distribution; influence of a wind speed and illumination on generation and energy storage battery life factors are comprehensively considered, and an optimization function taking lowest system generation cost as a target and a load power loss rate as a constraint condition is established; twice optimization is carried out by employing an improved harmony search algorithm, a needed wind generation power value, a photovoltaic generation power value and a theoretical energy storage battery capacity value are acquired through first optimization; the wind generation power value is adjusted to be a rated wind generator power existing in a market, and the target function is re-optimized to acquire the photovoltaic generation power and a final energy storage battery capacity target value which satisfy the constraint condition. Through the wind and photovoltaic complementation generation system optimization configuration method, a convergence speed in a solution process can be accelerated, secondary optimization satisfies practical demands, the optimization result is that power supply reliability can be improved, and system establishment cost can be reduced.

Description

A kind of wind and solar hybrid generating system Optimal Configuration Method

Technical field

The present invention relates to a kind of wind and solar hybrid generating system Optimal Configuration Method.

Background technology

Wind and solar hybrid generating system solves a kind of means that supply of electric power is carried out in the normal grid remote districts that can not arrive.Because wind and solar hybrid generating system does not have bulk power grid as the support of reserve capacity, the reliability service of the random fluctuation meeting influential system of wind speed and illumination.If adopt larger energy storage reserve capacity to meet power supply reliability, wind and solar hybrid generating system construction cost and maintenance cost will be increased.Wind energy and solar power generation have complementarity, and the reasonable disposition of carrying out wind and solar hybrid generating system can meet load supplying demand, improve power supply reliability, reduces cost of electricity-generating simultaneously, and then realizes the optimizing operation of wind and solar hybrid generating system better.

During wind and solar hybrid generating system is distributed rationally, rational Optimal Allocation Model and method for solving affect the accuracy of result.In existing research, using wind speed, light intensity in evaluation time section as invariant process, do not consider the correlation between wind speed illumination, modeling difficulty reduce but introduce larger error.When adopting probability density function to describe the statistical property model of wind power generation and photovoltaic generation, the probability density function of wind speed and illumination affects the accuracy of wind power generation and photovoltaic power generation power prediction.Do not have Unified Form to the description of wind speed and intensity of illumination probability distribution at present, wind speed commonly uses the descriptions such as Weibull distribution, gamma distribution, logarithm normal distribution; Intensity of illumination is commonly used beta distribution and is described.In fact the probability distribution of wind speed and intensity of illumination has difference, and Weibull distribution mainly describes the situation of annual mean wind speed.Existing document by wind-powered electricity generation, both photovoltaic generations exert oneself by obedience independent distribution process, by the wind speed under frequency statistics form calculus local condition, illumination joint probability distribution.The probability density function of wind speed, illumination not necessarily accurately meets specific exemplary distribution, therefore has certain limitation based on Parametric test research probability distribution.When setting up the target function that wind and solar hybrid generating system is distributed rationally, although have consider the factors such as depreciable cost, have ignored the correlation of wind speed and illumination.Some employings improves Lagrangian Arithmetic and is optimized wind and solar hybrid generating system, but chooses storage battery and have subjectivity.The modeling of some employing multiple targets, but when multiple target is changed into single goal, each target weighted value distribution has larger subjectivity.When solving objective optimization model, genetic algorithm is utilized to there is the problems such as convergence rate is slow, iterations is many.Particle cluster algorithm has that structure is simple, fast convergence rate, to advantages such as target function requirement are few, but also there is " precocity " problem, is easily absorbed in locally optimal solution.Chaotic optimization algorithm has the feature of randomness, ergodic and inherent law, but arithmetic accuracy is relevant with the size in optimizing space with the complexity of optimizing function.The selection of ant group algorithm parameter is comparatively large on optimum results impact, easily causes result non-optimal solution.There is the problems such as be easily absorbed in locally optimal solution and speed of searching optimization is slower in bacterial foraging algorithm.

Summary of the invention

The object of the invention is to the defect overcoming the existence of above-mentioned technology, there is provided a kind of wind and solar hybrid generating system to distribute the method for design rationally, reach the object improving renewable energy power generation utilization ratio, reduce wind and solar hybrid generating system construction cost and maintenance cost and enhancing system reliability.The present invention is by analyzing the correlation of exerting oneself between wind power generation and photovoltaic generation, and sign wind power generation and photovoltaic generation combine the complementary relationship of exerting oneself.According to wind power generation and photovoltaic generation characteristic, set up the model that wind light mutual complementing power generation is exerted oneself.Consider wind speed and illumination to factors such as the impact of generating and energy storage battery life-spans, set up with systems generate electricity cost the minimum majorized function being target and be constraints with load dead electricity rate, with improvement harmonic search algorithm to objective function optimization, on the basis of first optimum results, consider that actual conditions are revised result, again target function is optimized again, finally, be met the system optimization result of constraints.

Its concrete technical scheme is:

1 wind light generation combines the calculating of probability distribution of exerting oneself

1.1 wind power generations and photovoltaic generation are exerted oneself marginal probability distribution

In order to obtain combining wind and light to generate electricity power stage under DIFFERENT METEOROLOGICAL CONDITIONS, the joint probability distribution under honourable acting in conjunction need be known, and joint probability distribution can be obtained by marginal probability distribution.Affect by Changes in weather, wind speed and intensity of illumination probability density contour curve form are not fixed, and the present invention adopts non-parametric estmation to analyze probability density, use Density Estimator to draw the marginal probability density that wind-driven generator and photovoltaic cell are exerted oneself.According to the actual power of wind speed, photometric data calculating wind-driven generator, solar panel, and be that benchmark is normalized with rated output.To go out power rate for stochastic variable p, its probability density function is f (p), then the Density Estimator of f (p) is:

$f_h\left(p\right)=\frac{1}{nh}\underset{j=1}{\overset{n}{\Σ}}K\left(\frac{p-p_j}{h}\right)=\frac{1}{nh}\underset{j=1}{\overset{n}{\Σ}}\frac{1}{\sqrt{2\mathrm{\π}}}\exp (-\frac{{(P-P_j)}^2}{2h^2})---\left(1\right)$

In formula, n is sample size; H is smoothing factor; K (x) is kernel function, chooses kernel function here and obeys standardized normal distribution.

Wind power generation is monthly gone out power rate P1 (P in 1 year _1,1, P _1,2... P _1,12) and photovoltaic generation monthly to go out power rate be P2 (P _2,1p _2,2... P _2,12) substitute into (1) formula, namely obtain the marginal probability density f that exerts oneself when wind power generation and photovoltaic generation independent role _wGand f (P1) _pV(P2).Respectively to f _wGand f (P1) _pV(P2) carry out integral operation, the marginal probability distribution of wind power generation and photovoltaic generation can be obtained.

1.2 wind power generations and photovoltaic generation are exerted oneself correlation

Sample space Ф={ (P is formed by the observed value of random vector (P1, P2) _1,1, P _2,1), (P _1,2, P _2,2) ... (P _1,12, P _2,12), get (P _{1, i}, P _{2, i}) and (P _{1, j}, P _{2, j}) and i ≠ j, if (P _{1, i}-P _{1, j}) (P _{2, i}-P _{2, j}) > 0, then (P _{1, i}, P _{2, i}) and (P _{1, j}, P _{2, j}) consistent, otherwise the two is inconsistent.Kendall rank correlation parametric method has good characteristic when studying nonlinear correlation problem, and the present invention adopts the correlation of Kendall rank correlation Parameter analysis wind power generation and photovoltaic generation.Kendall rank correlation parameter τ is the difference of probability that the measured value chosen in sample is consistent and inconsistent probability:

τ＝P{(P _1i-P _1j)·(P _2j-p _2j)＞0}-P{(P _1i-p _1j)·(P _2i-p _2j)＜0}(2)

In formula, τ scope is [-1,1], the probability that P presentation of events occurs.

The joint probability distribution that 1.3 wind power generations and photovoltaic generation are exerted oneself

Wind power generation and photovoltaic generation are exerted oneself and are had negative correlation, can solve non-linear stochastic variable probability distribution by Copula function.Conventional Copula function is G-Copula, Clay-Copula and FrankCopula respectively, and wherein FrankCopula function can describe the negative correlation characteristic between non-thread sexual type variable.Here the joint probability distribution using Copula function to solve wind power generation and photovoltaic generation to exert oneself, concrete mathematical formulae is as follows:

$F(P_1,P_2)=-\frac{1}{\mathrm{\θ}}\ln \[1+\frac{(e^{-\mathrm{\θ}\·F_{WG}\left(P_1\right)}-1)\·(e^{-\mathrm{\θ}\·F_{PV}\left(P_2\right)}-1)}{e^{-\mathrm{\θ}}-1}\]---\left(3\right)$

In formula, relevant parameter θ and Kendall rank correlation parameters relationship as follows:

$\mathrm{\τ}=1+\frac{4}{\mathrm{\θ}}\[\frac{1}{\mathrm{\θ}}{\∫}_0^{\mathrm{\θ}}\frac{t}{e^{\′}-1}dt-1\]---\left(4\right)$

2 wind light mutual complementing power generations are exerted oneself model

2.1 wind power generation models

The energy output by wind machine output characteristic impact of wind-driven generator.Wind driven generator output power P _wGthe relational expression changed with wind speed V is as follows:

Wind energy conversion system energy output computing formula is as follows:

In formula, h _iit is the hourage corresponding to wind speed V.

2.2 photovoltaic generation models

The power stage of solar panel depends on area and the conversion efficiency of intensity of illumination, cell panel, therefore, provides following computational methods:

P _PV＝G·A·η(7)

In formula, P _pVfor photovoltaic cell real output, G is intensity of illumination (kW/m ²), A is photovoltaic battery panel area (m ²), η is the efficiency of photovoltaic cell.

Photovoltaic cell energy output every day E _pVfor

E _PV＝P _PV·h _d(8)

Wherein, h _daverage every day light application time.

2.3 battery model

In wind and solar hybrid generating system, storage battery is in charging and discharging two states, and accumulators store electricity relation is as follows:

C _bm(1-DOD)≤C _b≤C _bm(9)

In formula, C _bfor accumulator electric-quantity, DOD is depth of discharge, C _bmfor the maximum charge capacity of storage battery.

When systems generate electricity amount is greater than load electricity consumption, storage battery is in charged state, and storage battery energy relational expression is as follows:

C _b(t)＝C _b(t-1)+[P _PV(t)+P _WG(t)-P _LOAD]·η _ch≤C _bm(10)

In formula, C _bt () is the electricity of t storage battery, η _chfor charge efficiency.

When systems generate electricity amount is less than load electricity consumption, storage battery is in discharge condition, and storage battery energy relational expression is as follows:

C _b(t)·η _dis＝C _b(t-1)+[P _LOAD(t)-P _PV(t)-P _WG(t)]≥C _bm(1-DOD)(11)

In formula, C _bt () is the electricity of t storage battery, η _disfor discharging efficiency.

3 objective optimizations

3.1 target function

The target that wind and solar hybrid generating system is distributed rationally is under proof load power supply reliability and the prerequisite in system useful life, and the integrated cost that the primary construction cost of system and later maintenance cost are formed is minimum.Its target function is as follows:

min{J(x)}＝min{J(P _PV，P _WG，P _BAT)}(12)

＝min{P _PV·C _PV+P _WG·C _WG+P _BAT·C _BAT+n·(P _PV·C _PVm+P _WG·C _WGM+P _BAT·C _BM)+C _IV+C _CL}

In formula, J (x) is integrated cost (system primary construction cost and maintenance cost common expense), and x is kilowatt number of part, X={P _pV, P _wG, P _bAT; P _pVfor meeting the photovoltaic cell capable of generating power gross power of loading demand, P _wGfor meeting the wind turbine power generation gross power of loading demand, P _bATfor the storage battery total capacity that load needs; C _wG, C _pV, C _bATbe respectively wind-driven generator, photovoltaic cell, battery cell's acquisition cost, C _wGM, C _pVM, C _bMbe respectively wind-driven generator, photovoltaic cell, battery cell's maintenance cost; N is service life, C _iV, C _cLfor inverter and controller cost.

Above-mentioned target function not only calculates system preliminary expenses but also adds the maintenance cost in service life, is more close to reality.

3.2 constraints

(1) photovoltaic power constraint

0≤P _PV≤P _PVmax(13)

Wherein, P _pVfor load needs photovoltaic cell capable of generating power gross power; P _pVmaxfor under calm condition, maximum power of photovoltaic cell and bearing power when solar cell is independently-powered.

(2) wind-driven generator power constraint

0≤P _WG≤P _WGmax(14)

Wherein, p _wGfor load needs wind turbine power generation gross power; P _wGmaxfor under no light condition, wind-driven generator maximum power and bearing power when wind-driven generator is independently-powered.

(3) storage battery heap(ed) capacity constraint

0≤P _BAT≤P _BATmax(15)

Wherein, P _bATfor load needs the total capacity of accumulators store electricity; P _bATmaxfor under calm, no light condition, the heap(ed) capacity of storage battery when storage battery is independently-powered.

(4) load dead electricity rate LOLP retrains

LOLP under electricity that system produces can not meet loading demand condition, the ratio of the electricity lacked and load aggregate demand, its expression formula is,

$\frac{P_{load}-\[(P_{PV}+P_{WG})\·p+\frac{P_{BAT}\·U}{1000\·t}\·(1-p)\]}{P_{load}}\≤{\mathrm{LOLP}}_{set}---\left(16\right)$

In formula, P _loadfor load gross power, p is that wind power generation combines with photovoltaic generation probability of exerting oneself, and U is the system busbar supply power voltage chosen, and t is calm the longest unglazed hourage.

3.3 improve harmonic search algorithm

Harmonic search algorithm to simulate in musical performance between musician mutual break-in until whole playing effect reaches the process of the most U.S. harmony.Although this algorithm has, versatility is good, thought principle is simple, with the advantage such as other algorithm associativities are strong, harmonic search algorithm also has the shortcoming being absorbed in local optimum, and the state modulator of algorithm has a great impact convergence rate.Fixed variable disturbance benchmark and these two parameters of disturbance bandwidth during the initialization of standard harmonic search algorithm, make the solution produced easily be absorbed in local optimum.The present invention adopts the harmonic search algorithm of improvement by variable disturbance benchmark and disturbance bandwidth along with iterations dynamic conditioning, thus global optimizing.

Harmonic search algorithm solution procedure is as follows:

(1) initialization harmony storehouse

The harmony storehouse that a series of initial solution of random generation is formed, its storage capacity is HMS.For ensureing that the solution produced has certain representativeness, and be uniformly distributed in the domain of definition, initial solution X _ijproduced by following formula,

X _ij＝LB _i+Rnd1×(UB _i-LB _i)(17)

In formula, X _i∈ [LB _i, UB _i], j ∈ (1,2......HMS), LB _i, UB _ibe respectively interval maximum and minimum value, Rnd1 is the random number being uniformly distributed generation between (0,1).

(2) generation of separating

Specify the pick ratio HMCR producing solution from harmony storehouse.Produce random number R nd2, when Rnd2 is less than HMCR, choose a solution at random from harmony storehouse, otherwise just produce new solution by formula (17), be designated as Ω here.Employing formula (18) produces primitive solution:

$x_i=\left\{\begin{array}{cc}{x}_i\&Element;(x_{i1},x_{i2},...x_{iHMS})& Rnd2<HMCR\\ x_i\&Element;\mathrm{\&Omega;}& Rnd2\&GreaterEqual;1-HMCR\end{array}\right.---\left(18\right)$

Micro-disturbance is carried out to primitive solution, adopts formula (19) to produce new explanation:

$x_i=\left\{\begin{array}{cc}{x}_i+Rnd3\&times;bw& Rnd3<PAR\\ x_i& Rnd3\&GreaterEqual;1-PAR\end{array}\right.---\left(19\right)$

In formula, random produce amount trimmed Rnd3 × bw, bw are disturbance bandwidth, Rnd3 is random number, and PAR is variable disturbance benchmark.Fix these two parameters of PAR and bw during the initialization of standard harmonic search algorithm, make PAR when comparing with random number R nd3, Rnd3 always compares with same PAR, can repeatedly produce identical amount trimmed, thus the solution produced does not have of overall importance.

The harmonic search algorithm improved is by PAR and bw along with iterations dynamic conditioning, and PAR is by formula (20) adjustment, and bw adjusts by formula (21),

PAR＝(PAR _max-PAR _min)/MItr×cItr+PAR _min(20)

In formula, PAR _maxand PAR _minfor the span of PAR.MItr is maximum iteration time, and cItr is current iteration number of times.

bw＝bw _max×exp(ln(bw _min/bw _max)/MItr×cItr)(21)

In formula, bw _minand bw _maxbe respectively maximum and the minimum value of bandwidth.

(3) data base is upgraded

New explanation substituted in target function, the desired value obtained compares with the desired value of separating in harmony storehouse, if the target function value of new explanation is better than the desired value of the poorest solution in harmony storehouse, then replaces the poorest solution in harmony storehouse with new explanation, otherwise does not replace.

(4) algorithm stops

When obtaining optimal solution or arriving maximum iteration time, stop computing.

Compared with prior art, beneficial effect of the present invention is: first, analyzes the correlation of wind power generation and photovoltaic generation, obtains the joint probability distribution of wind speed and illumination; Secondly, set up with the minimum majorized function being target and be constraints with load dead electricity rate of systems generate electricity cost; Again, adopt the harmonic search algorithm improved to objective function optimization, final goal value is obtained by twice optimizing process, optimize the wind power generation acc power of the demand that obtains, photovoltaic generation power and energy-storage battery capability value for the first time, then wind power generation power is rounded the wind power generation acc power into market exists, target function is being optimized again, finally, is being met the system optimization result of constraints.

Accompanying drawing explanation

Fig. 1 illumination probability density

Fig. 2 wind speed probability density

Fig. 3 wind power generation and photovoltaic generation are exerted oneself joint probability distribution

Fig. 4 objective function optimization result

Fig. 5 systems generate electricity amount and load power consumption relation

Embodiment

The technological means realized to make the present invention, creation characteristic, reaching object and effect is easy to understand, below in conjunction with accompanying drawing and instantiation, setting forth the present invention further.

Choose Dalinuoer lake area, Keshiketeng Banner, Chifeng and carry out family wind and solar hybrid generating system profile instance calculating, this area is positioned at north latitude 43 °, east longitude 117 °.According to the data that local meteorological station provides, in conjunction with the data information of NASA website, according to the actual power of wind speed, photometric data calculating wind-driven generator, solar panel, and be that benchmark is normalized with rated output.Going out power rate wind power generation each moon is P1=(0.072,0.046,0.040,0.043,0.061,0.005,0.022,0.006,0.014,0.021,0.065,0.094), going out power rate PV assembly each moon is P2=(0.3,0.377,0.512,0.688,0.773,0.80,0.692,0.615,0.592,0.515,0.358,0.277).

1 wind power generation and photovoltaic generation are exerted oneself correlation

Wind power generation and photovoltaic generation each moon are gone out power rate data and brings formula (1) into, the wind speed marginal probability density f shown in Fig. 1 can be obtained _wG(P1) the illumination marginal probability density f and shown in Fig. 2 _pV(P2).And then integration can obtain the marginal probability distribution F of wind speed and illumination _wGand F (P1) _pV(P2).

Correlation is estimated

Wind power generation and photovoltaic generation each moon go out power rate set Ф={ (0.3,0.072), (0.377,0.046), (0.512,0.040), (0.688,0.043), (0.773,0.061), (0.8,0.005), (0.692,0.022), (0.615,0.006), (0.515,0.014), (0.358,0.021), (0.277,0.065), (0.5,0.094) sample space of random vector (P1, P2) } is formed.It is 0.242 that calculating wind power generation changes consistent probability with photovoltaic generation, and inconsistent probability is 0.758, utilizes formula (2) to obtain coefficient correlation τ=-0.516, and this illustrates that wind power generation and photovoltaic generation have negative correlation characteristic.

2 wind power generations and photovoltaic generation are exerted oneself joint probability distribution

After known coefficient correlation τ, parameter θ=-6.04 in Copula function are calculated according to formula (4), thus wind power generation and photovoltaic generation can be calculated by formula (3) and to exert oneself joint probability distribution, the joint probability distribution that wind power generation and photovoltaic generation are exerted oneself is as shown in Figure 3.

3 objective optimizations

Household electrical appliance kind in man of resident family by inquiry, the daily power consumption of statistics and each moon power consumption.According to adding up the customer charge curve obtained, choose the load that maximum power is 5kW.According to 1 year continuous overcast and rainy days and air speed data in meteorological data; draw the time under continuous weak wind and non-illuminated conditions; under load dead electricity rate 1% condition; the battery capacity needed under calculating extreme condition; consider the safe handling condition of protection storage battery; the maximum depth of discharge of storage battery is set as 80%, draws as calculated, maximum short of electricity amount 10kWh.If the DC bus-bar voltage of wind and solar hybrid generating system is 24V, 12V storage battery connection in series-parallel form is adopted to expand capacity.According to market survey, wind power generation and photovoltaic generation unit acquisition cost and maintenance cost as shown in table 1.

Table 1 wind power generation and photovoltaic generation unit acquisition cost and maintenance cost

Utilize the harmonic search algorithm improved to carry out system optimization configuration, in optimizing process, each parameter is respectively, HMS=5, HMCR=0.9999, PAR _min=0.4, PAR _max=0.9.Objective function optimization process as shown in Figure 4.After first time is optimized, the total kilowatt number of wind-driven generator is 3.397kW, but market does not have the wind-driven generator of this power model, therefore wind power generation acc power is adjusted to 3kW wind-driven generator common on market, carry out second time optimization, obtain the size of photovoltaic power kilowatt number and battery capacity.It is 1.997kW that second time optimizes rear photovoltaic power, can be similar to 2kW.Battery capacity is 500AH.Twice optimum results contrast is as shown in table 2.

Table 2 optimum results contrasts

In order to verify the correctness of optimum results, the wind and solar hybrid generating system of calculation optimization configuration meets load supplying rate situation.Fig. 5 is the system power supply curve and load curve that calculate.Make a general survey of annual generating curve, illumination in summer is strong, and solar power generation amount is many, and 6-9 month wind power generation amount is less, and in a word, solar power generation and wind power generation have good complementarity., simultaneously wind speed and illumination maximum with power consumption is not carry out energy output calculating very sufficient month to compare, and show that load short of electricity rate LOLP is 0.97%, is less than design objective 1%, can meets power supply reliability requirement.

The above, be only best mode for carrying out the invention, and any those skilled in the art of being familiar with disclose in technical scope in the present invention, and the simple change of the technical scheme that can obtain apparently or equivalence are replaced and all fallen within the scope of protection of the present invention.

Claims (5)

1. a wind and solar hybrid generating system Optimal Configuration Method, concrete enforcement comprises the following steps:

(1) the electric calculating of exerting oneself of associating is generated to wind power generation and photovoltaic.Photovoltaic generation is exerted oneself marginal probability density to adopt nonparametric probability to draw; Nonparametric probability is adopted to draw wind power generation output marginal probability density; Calculate wind light generation correlation; Utilize Copula function to calculate wind light generation to exert oneself joint probability distribution; Calculate wind power generation and photovoltaic generation to combine and exert oneself;

(2) consider wind speed and illumination to factors such as the impact of generating and energy storage battery life-spans, set up with systems generate electricity cost minimum for target and the majorized function that is constraints with load dead electricity rate;

(3) adopt the harmonic search algorithm improved to objective function optimization, first suboptimization obtains the wind power generation acc power of demand, photovoltaic generation power and energy storage battery Capacity Theory value;

(4) wind power generation power definite value is adjusted to the wind power generation acc power that market exists, again target function is optimized again, be met photovoltaic generation power and the energy storage battery capacity final goal value of constraints.

2. a kind of wind and solar hybrid generating system Optimal Configuration Method according to claim 1, it is characterized in that: to calculate premised on wind light generation correlation, adopt nonparametric probability and Copula function to combine to exert oneself to wind power generation and photovoltaic generation and calculate.

3. a kind of wind and solar hybrid generating system Optimal Configuration Method according to claim 1, it is characterized in that: consider wind speed and illumination to factors such as the impact of generating and energy storage battery life-spans, set up with systems generate electricity cost minimum for target and the majorized function that is constraints with load dead electricity rate.

4. a kind of wind and solar hybrid generating system Optimal Configuration Method according to claim 1, is characterized in that: adopt the harmonic search algorithm improved to objective function optimization.

5. a kind of wind and solar hybrid generating system Optimal Configuration Method according to claim 1, is characterized in that: adopt twice optimal way to target function, obtain final result.It is theoretical that first suboptimization obtains the wind power generation acc power of demand, photovoltaic generation power and energy storage battery capability value; Wind power generation power is adjusted to the wind power generation acc power that market exists, again target function is optimized again, be met photovoltaic generation power and the energy storage battery capacity final goal value of constraints.