CN104616067A - Wind power storage power generation control method and system considering grid purchase price and genetic algorithm optimization - Google Patents

Abstract

The invention discloses a wind power storage power generation control method and system considering grid purchase price and genetic algorithm optimization. The wind power storage power generation control method includes steps that setting parameters of a genetic algorithm; generating all the individuals of an initial population; calculating the adaptive value of each individual; generating a new population; carrying out crossover operation on each individual of the successfully paired individuals of the new population; carrying out mutation operation on each of the individuals performed with the crossover operation; regulating all the individuals according to the upper and lower limit constraint of the power of an energy storage system and the upper and lower limit constraint of the charge state; judging whether the new population meets an iteration times requirement or a convergence requirement, if so, outputting the new population, otherwise, continuing to iterate; selecting the optimal individual and the global optimum from the new population, and outputting the charge/discharge power of the energy storage system at each time. The system comprises a population generating module, a calculating module, a crossover module, a mutation module and an execution module. The wind power storage power generation control method and system considering the grid purchase price and genetic algorithm optimization reduce the operation cost while restraining a large wind electricity ramp rate.

Description

Consider that the wind of rate for incorporation into the power network and genetic algorithm optimization stores up electricity-generating control method and system

Technical field

The invention belongs to accumulator system discharge and recharge technology, be specifically related to a kind ofly consider that the wind of rate for incorporation into the power network and genetic algorithm optimization stores up electricity-generating control method and system, in particular to wind storage system economic load dispatching and the Optimal Control Problem of considering operating cost and economic benefit, and solve emphatically the control problem of accumulator system discharge and recharge.

Background technology

Along with the development of wind power technology, the large-scale grid connection of wind-powered electricity generation, its fluctuation problem of exerting oneself is day by day serious.Because wind-powered electricity generation belongs to regenerative resource, its size of exerting oneself is subject to the impact of the factor such as weather, landform, causes the uncertainty that it is exerted oneself, and the size of exerting oneself can not keep constant always, and be change at any time in a lot of degree, there is very strong undulatory property and intermittence.After wind power integration electrical network, its undulatory property and intermittence all can cause adverse effect to the stability of operation of power networks.Therefore, in practice, the grid-connected of wind-powered electricity generation is limited by very large, and is unfavorable for the development of wind-powered electricity generation.Battery energy storage system has response fast and carries out the ability of discharge and recharge, can follow the tracks of exerting oneself of wind-powered electricity generation constantly, therefore, the battery energy storage system of wind-powered electricity generation outfit certain capacity is contributed to the suppression of wind-powered electricity generation climbing rate, reduces the impact of wind-powered electricity generation fluctuation on electrical network.

The suppression of wind power output fluctuation is related to the problem of wind-electricity integration income.Wind energy turbine set provides electric energy can obtain certain income to electrical network, but electrical network is in order to ensure the operation of himself stability and safety, can limit to some extent the climbing rate of wind power output, when wind power output exceedes the restriction of grid-connected climbing rate, certain economy punishment can be carried out to wind energy turbine set, or do not allow wind-electricity integration completely, now will cause loss economically to the operation of wind energy turbine set self.Therefore wind energy turbine set is in order to ensure the online of its generated energy, realize the maximization of economic benefit, can by being equipped with the battery energy storage system of certain capacity, according to the situation of wind power output, adopt the discharge and recharge operation of battery energy storage system, global optimization is carried out to climbing rate, to meet the requirement of electrical network to wind-powered electricity generation climbing rate, and then realizes the economic benefit of wind energy turbine set sale of electricity.In practice, battery energy storage system is that discharge and recharge operates by fixing discharge and recharge working value, and therefore, the Optimal Control Problem of battery energy storage system discharge and recharge operation is the problem of discretize.In addition, under considering the running environment of intelligent grid and electricity market, based on rate for incorporation into the power network, various Cost and benefit etc., how to process Large Copacity accumulator system economic load dispatching and Optimal Control Problem is also key core problem urgently to be resolved hurrily, but correlative study achievement is less at present.

In addition, due to the discharge and recharge optimization problem suppressing wind-powered electricity generation climbing rate problem to relate to multi-period battery energy storage system discretize, belong to the problem of higher-dimension discontinuum, traditional optimized algorithm cannot solve problems, therefore in the present invention, genetic algorithm is improved, to solve this problem.

Summary of the invention

In order to overcome the above-mentioned defect of prior art, an object of the present invention is to propose a kind ofly to consider that the wind of rate for incorporation into the power network and genetic algorithm optimization stores up electricity-generating control method.The present invention mainly adopts improved genetic algorithms law technology, consider online electrical network and operating cost, using the charge-discharge electric power value of battery energy storage system as gene, the discharge and recharge of battery energy storage system is optimized, reduces gene, accelerate search speed, to realize reasonable utilization accumulator system, reduce wind-powered electricity generation climbing rate, control and the maximized object of on-road efficiency with the generating optimization realizing wind storing cogeneration system, there is very large practical value.

In order to realize foregoing invention object, method of the present invention is achieved through the following technical solutions:

Consider that the wind of rate for incorporation into the power network and genetic algorithm optimization stores up an electricity-generating control method, it comprises the following steps:

The parameters such as A, the correlation parameter that genetic algorithm is set and rate for incorporation into the power network.

B, the charging and discharging state fixed value sequence set according to battery energy storage system, with it for gene, generate the initial population of genetic algorithm.

The adaptive value of each individuality in C, calculating population.

D, determine to select by the method for roulette and copy follow-on individuality.

E, the new individuality random pair between two will obtained in step D, and determine whether according to crossover probability the random interlace operation carrying out battery energy storage system charge-discharge electric power, obtain new individuality.

The new individuality obtained in F, step e, according to mutation probability, carries out the random variation operation of battery energy storage system charge-discharge electric power, obtains new individuality.

G, according to the constraint of the state-of-charge of battery energy storage system, individuality to be adjusted, make it not out-of-limit.

Whether the new population that H, judgement obtain has met iteration requirement, if met the demands, then and iteration stopping, otherwise forward step C continuation iteration to.

I, according to the population obtained in step G, find the optimum individual in this population, Calculation Estimation suppresses the objective function of wind-powered electricity generation climbing rate, climbing rate index, exports the discharge and recharge operating result of battery energy storage system.

Further, in steps A, the parameters needed for genetic algorithm is set, comprises crossover probability, mutation probability, iterations, Population Size.

Further, in step B, according to the charging and discharging state fixed value sequence of battery energy storage system setting, using it as the gene of population, the population of stochastic generation genetic algorithm, wherein the individuality of population is the charge-discharge electric power of day part battery energy storage system.

Further, in step C, to the individuality obtained in step B, calculate the functional value of each individuality according to the objective function of wind storage commingled system generating optimization, calculate the difference of each individual functional value and wherein minimum function value, and in this, as the adaptive value of individuality.

Further, in step D, in calculation procedure C obtain the cumulative probability of individual fitness, and individual and it is copied according to roulette method choice, obtain new population.

Further, in step e, by the population at individual random pair obtained in step D.Interlace operation is carried out to every a pair individuality: generate the random number of 0 to 1, the relatively size of itself and crossover probability, if be less than crossover probability, carry out interlace operation, during operation, the exchange of gene is carried out in random selected genes position, namely the value of a certain moment charge-discharge electric power of Stochastic choice battery energy storage system exchanges, complete interlace operation, thus obtain new population.

Further, in step F, to in step e obtain new population each individuality carry out mutation operation one by one: generate the random number of 0 to 1, the relatively size of itself and mutation probability, if be less than mutation probability, carry out mutation operation, during operation, the variation of gene is carried out in random selected genes position, namely upgrades the value of battery energy storage system charge-discharge electric power, cause mutation operation, thus obtain new population.

Further, in step G, the state-of-charge according to battery energy storage system retrains each gene in individuality, and namely the charge-discharge electric power of energy-storage battery adjusts, and avoids it out-of-limit.

Further, in step H, judge whether the requirement reaching iterations or meet convergence, require if reach iteration, stop iteration, export the population after optimizing, require if do not reach iteration, forward in step C and continue iteration.

Further, in step I, according to the population after the optimization obtained in step G, find out individuality optimum in this population, calculate its target function value, the climbing rate index after optimization, and the discharge and recharge generating optimization result exporting battery energy storage system.

Another object of the present invention is to propose a kind ofly consider that the wind of rate for incorporation into the power network and genetic algorithm optimization stores up power-generating control system, it comprises:

Population generation module, for arranging the parameter of genetic algorithm, and generates initial population;

Computing module, for calculating adaptive value and the cumulative probability thereof of each individuality in initial population, carries out selection replicate run to each individuality, generates new population;

Cross module, for the random pair between two of the individuality in new population, and carries out interlace operation to the individuality of often pair of successful matching;

Variation module, for carrying out mutation operation to carrying out each individuality after interlace operation; With

Execution module, for judging whether new population meets iterations requirement or convergent requirement, until export new population after meeting arbitrary requirement, and find out the corresponding individual position of global optimum wherein, the position of described individuality is the charge-discharge electric power of the corresponding accumulator system of this individuality.

Compared with prior art, the present invention has following beneficial effect:

1, the invention provides a kind of energy storage multi objective control method and the system thereof of considering rate for incorporation into the power network and genetic algorithm, the present invention considers the inhibiting effect of battery energy storage system to wind-powered electricity generation climbing rate, have employed the discharge and recharge operation of a kind of Revised genetic algorithum to battery energy storage system to be optimized, by adding the out-of-limit penalty of wind-powered electricity generation climbing rate, storage energy operation cost, wind storing cogeneration income etc. in objective function, limit the problem that wind-powered electricity generation climbing rate is out-of-limit well, achieve while suppressing wind-powered electricity generation climbing rate excessive, obtain the object of economical operation benefit.In use genetic algorithm optimization process, using calculating the difference of each individual goal functional value and all individual goal functional values as its adaptive value, the fitness calculating each individuality in population can be found better.

2, the control method proposed in the present invention is applied, not only consider that the control overflow of satisfied climbing rate constraint of exerting oneself is combined in wind storage, and take into account rate for incorporation into the power network, energy storage discharge and recharge cost, wind power output climbing rate out-of-limit time the factor such as rejection penalty, wind storing cogeneration system grid connection gene-ration revenue, and pass through based on genetic algorithm optimizing, achieve the wind storing cogeneration systematic economy optimal control object that wind storing cogeneration net proceeds is maximum, effectively improve the control efficiency of accumulator system.

Accompanying drawing explanation

Fig. 1 is the process flow diagram that the present invention considers the accumulator system multiobjective optimization control method of rate for incorporation into the power network;

Fig. 2 is that wind storage jointly controls optimum results curve map;

Fig. 3 is that wind power output climbing rate optimizes forward and backward comparison diagram.

Embodiment

Below in conjunction with accompanying drawing, preferred embodiment is elaborated.It is emphasized that following explanation is only exemplary, instead of in order to limit the scope of the invention and apply.The applied environment of this example is wind storage association system, and comprising wind energy turbine set and accumulator system, in this example, accumulator system adopts battery energy storage system.

Consideration rate for incorporation into the power network in this example and the wind of genetic algorithm optimization store up power-generating control system and comprise:

A, population generation module, for arranging the parameter of genetic algorithm, and generate all individualities in initial population.Population generation module may further include:

A1, optimum configurations submodule, for arranging the parameter of genetic algorithm according to actual conditions;

A2, the first calculating sub module, for according to accumulator system charge-discharge electric power state, all individualities in stochastic generation initial population.

B, computing module, for calculating adaptive value and the cumulative probability thereof of each individuality in initial population, carry out selection replicate run to each individuality, generates new population.Computing module may further include:

B1, the second calculating sub module, for calculate wind storage commingled system out-of-limit to the wind power output value of electrical network, wind power output climbing rate and wind power output climbing rate time penalty term, and determine the adaptive value of each individuality further by the objective function F of wind storage commingled system generating optimization;

B2, cumulative probability calculating sub module, for asking for the cumulative probability of each individual fitness according to adaptive value;

B3, replicon module, for the comparative result according to cumulative probability and random number, adjust each individuality and generate new population.

C, Cross module, for the random pair between two of the individuality in new population, and carry out interlace operation to the individuality of often pair of successful matching.Cross module may further include:

C1, the first comparison sub-module, for by comparing crossover probability p _cwith random number r ₃size, determine whether carrying out interlace operation to the individuality of a pair successful matching every in population, described random number is the random number of 0 to 1 of stochastic generation;

C2, interlace operation submodule, if desired carry out interlace operation, then first determine to carry out interlace operation to the accumulator system charge-discharge electric power in which moment; And when interlace operation, the power in the determined moment of the individuality of successful matching is exchanged, to complete interlace operation.

D, variation module, for carrying out mutation operation to carrying out each individuality after interlace operation.Variation module may further include:

D1, the second comparison sub-module, for by comparing mutation probability p _mwith random number r ₅size, determine whether carrying out mutation operation, described r to particle each in population ₅for the random number of 0 to 1 of stochastic generation;

D1, mutation operation submodule, if desired carry out mutation operation, then first determine to carry out mutation operation to the accumulator system charge-discharge electric power in which moment; And when mutation operation, according to the maximal value of accumulator system charge and discharge power to the accumulator system charge-discharge electric power value again in this moment.

E, execution module, for judging whether new population meets iterations requirement or convergent requirement, until export new population after meeting arbitrary requirement, and find out the corresponding individual position of global optimum wherein, the position of described individuality is the charge-discharge electric power of the corresponding accumulator system of this individuality.Execution module may further include:

E1, judge module, for judging whether new population meets iterations requirement or convergent requirement, described iterations requires to arrange according to actual conditions, and described convergent requirement is that all ideal adaptations are worth the difference of maximal and minmal value to be less than a minimum number being greater than zero;

E2, result output module, for selecting global optimum individuality from the new population meeting iterations requirement or convergent requirement, and recalculate the adaptive value of this individuality by computing module, and export the accumulator system charge-discharge electric power corresponding to this individuality.

Objective function

When suppressing wind power output climbing rate excessive, target is turned to so that wind-powered electricity generation electricity volume is maximum, consider the cost that battery energy storage system operation (i.e. discharge and recharge operation) brings and penalty term when wind power output climbing rate is out-of-limit simultaneously, obtain objective function as follows:

$\max F_i^j={\mathrm{\Σ}}_{t=1}^T[c_{\mathrm{grid}}\·P_{\mathrm{grid},i,t}^j\·\mathrm{\Δt}-c_{\mathrm{bess}}\·|P_{\mathrm{bess},i,t}^j|\·\mathrm{\Δt}-M\·P_{\mathrm{punish},i,t}^j\·\mathrm{\Δt}]---\left(1\right)$

$P_{\mathrm{grid},i,t}^j=P_{\mathrm{wind},t}+P_{\mathrm{bess},i,t}^j---\left(2\right)$

$P_{\mathrm{punish},i,t}^j=\left\{\begin{array}{c}{P}_{\mathrm{ramp},i,t}^j-P_{\mathrm{stad}},\mathrm{if}{P}_{\mathrm{ramp},i,t}^j-P_{\mathrm{stad}}>0\\ 0,\mathrm{if}{P}_{\mathrm{ramp},i,t}^j-P_{\mathrm{stad}}\≤0\end{array}\right.---\left(3\right)$

In formula (1)-(3), i=1,2 ..., N, N represent the Individual Size in population, represent jth generation i-th individuality target function value, T be optimize time hop count, c _gridfor rate for incorporation into the power network, can arrange according to actual conditions, for according to jth generation i-th individuality t the power that the wind storage commingled system calculated provides to electrical network, P _{wind, t}for the wind power output of t, c _bessfor the unit operation cost of battery energy storage system, for the charge-discharge electric power value of battery energy storage system t, wherein negative value represents charging, on the occasion of expression electric discharge, 0 represents that battery energy storage system is in floating charge state, its charge and discharge process all can bring corresponding cost, therefore in objective function, absolute value is asked for it, M is wind-powered electricity generation climbing rate out-of-limit penalty term coefficient, its value wants the unit operation cost of long-range little rate for incorporation into the power network and battery energy storage system, penalty term coefficient can be set to 100 yuan/MWh for according to jth generation i-th individuality t penalty term when the wind power output climbing rate calculated the is out-of-limit difference of value (the wind power output climbing rate be above standard), for according to jth generation i-th individuality t the wind power output climbing rate calculated, P _stadfor according to jth generation i-th individuality t calculate wind power output climbing rate standard value, Δ t is the interval that wind power output is measured.

Wind power output climbing rate is defined as follows:

$P_{\mathrm{ramp},i,t}^j=\left\{\begin{array}{c}\left|{\mathrm{\&Delta;P}}_{\max +,i,t}^j\right|,\mathrm{if}\left|{\mathrm{\&Delta;P}}_{\max +,i,t}^j\right|\&GreaterEqual;\left|{\mathrm{\&Delta;P}}_{\max -,i,t}^j\right|\\ \left|{\mathrm{\&Delta;P}}_{\max -,i,t}^j\right|,\mathrm{if}\left|{\mathrm{\&Delta;P}}_{\max +,i,t}^j\right|<\left|{\mathrm{\&Delta;P}}_{\max -,i,t}^j\right|\end{array}\right.---\left(4\right)$

In formula, for according to jth generation i-th individuality t the difference of maximal value in the wind power output value calculated is interval with [t, t-K], for according to jth generation i-th individuality t the difference of minimum value in the wind power output value calculated is interval with [t, t-K], K is the time period of set investigation when suppressing wind power output climbing rate.

Constraint condition

The constraint condition of battery energy storage system power bound is:

P _min≤P _bess，t≤P _max(5)

In formula, P _minfor negative value, represent the lower limit of charge power, P _maxfor on the occasion of, represent the upper limit of discharge power.

The constraint condition of battery energy storage system state-of-charge bound is:

SOC _min≤SOC _t≤SOC _max(6)

In formula, SOC _min, SOC _maxbe respectively minimum, the maximum state-of-charge that battery energy storage system allows, SOC _tfor the state-of-charge of t battery energy storage system.

The state-of-charge of battery energy storage system is calculated as follows:

${\mathrm{SOC}}_t={\mathrm{SOC}}_{t-1}-\frac{P_{\mathrm{bess},t}*\mathrm{\&Delta;t}}{C_{\mathrm{bess}}}\&times;100\%---\left(7\right)$

In formula, SOC _t, SOC _t-1for the state-of-charge of t, t-1 moment battery energy storage system, C _bessfor the rated capacity of battery energy storage system, P _{bess, t}for the charge-discharge electric power of t battery energy storage system.

In view of in practice, the charge-discharge electric power of battery energy storage system is the value of a series of fixing discretize, namely

$P_{\mathrm{bess},i}^j=[{P_{\mathrm{bess},i,1}}^j,{P_{\mathrm{bess},i,2}}^j,...,{P_{\mathrm{bess},i,t}}^j,...,,{P_{\mathrm{bess},i,T}}^j]---\left(8\right)$

In formula, i=1,2 ..., N, N represent the Individual Size in population, and T is the moment number optimized, P _{bess, i, 1} ^jrepresent the charge-discharge electric power of t period accumulator system, altogether T period. for i-th individuality in initial population, be the charge-discharge electric power value of t battery energy storage system. value from table 1, therefore can not violate charge-discharge electric power constraint.

As shown in Figure 1, the consideration rate for incorporation into the power network that proposes of this example and the wind of genetic algorithm optimization store up electricity-generating control method and comprise the steps:

Steps A: the correlation parameter arranging genetic algorithm, comprises crossover probability p _c, value can be 0.9, mutation probability p _m, value can be 0.1, Population Size N, value can be 40 and iterations C, value can be 500.

Step B: in conjunction with the jth of battery energy storage system for i-th individual charging and discharging state sequence in population $P_{\mathrm{bess},i}^j=[{P_{\mathrm{bess},i,1}}^j,{P_{\mathrm{bess},i,2}}^j,...,{P_{\mathrm{bess},i,t}}^j,...,{P_{\mathrm{bess},i,T}}^j],$ Each individuality in stochastic generation initial population (i.e. 1st generation population) namely the gene of each gene point of each individuality is determined at random, the charge-discharge electric power of the gene representation day part battery energy storage system of each gene point.

Step C: calculate wind storage commingled system each moment according to formula (2) and go out force value (namely wind stores up the power that provides to electrical network of commingled system each moment) P to electrical network _grid, calculate climbing rate according to formula (4), calculate penalty term according to formula (3), calculate each individual corresponding target function value according to formula (1), calculate the adaptive value of each individuality according to following formula:

${f_i}^j=F_i^j-F_{\min }^j---\left(9\right)$

In formula, represent i-th individual target function value, represent the minimum value of all individual goal functional values in population.

Step D: the cumulative probability calculating each individual fitness:

$p_i=\left\{\begin{array}{c}{f_i}^j/{\mathrm{\&Sigma;}}_{i=1}^N{f_i}^j,\mathrm{ifi}=1\\ p_{(i-1)}+{f_i}^j/{\mathrm{\&Sigma;}}_{i=1}^N{f_i}^j,\mathrm{ifi}\&GreaterEqual;2\end{array}\right.---\left(10\right)$

In formula, p _ibe i-th individual cumulative probability, f _ibe i-th individual adaptive value.

According to following formula, selection replicate run is carried out to all individualities, generates new population:

$P_{\mathrm{bess},i}^j=\left\{\begin{array}{c}{P}_{\mathrm{bess},1}^{j-1},\mathrm{if}{r}_2\&le;p_1\\ P_{\mathrm{bess},k}^{j-1},\mathrm{if}{p}_k<r_2\&le;p_{k+1}\end{array}\right.---\left(11\right)$

In formula, represent i-th individuality in jth generation, r ₂it is the random number between 0 to 1.

Step e: carry out interlace operation to copying the new population obtained, new population is random pair between two.

For first, interlace operation whether judgement is carried out to individuality, generate the random number r of 0 to 1 ₃if, crossover probability p _c≤ r ₃, then a certain gene location P in Stochastic choice individuality _{bess, i, t} ^jcarry out interlace operation, P _{bess, i, t} ^jrepresent the energy storage charge-discharge electric power of i-th individual t in jth generation, the moment, t was for determine at random; If crossover probability p _c> r ₃, then interlace operation is not carried out.All aforesaid operations is carried out to all pairing individualities, to complete whole crossover process.

Determine a certain gene location in individuality by following formula, namely determine to carry out interlace operation to the accumulator system charge-discharge electric power of a certain moment t:

t＝[T*r ₄]

In formula, T is hop count during simulation (when optimizing hop count); r ₄for the random number of 0 to 1 of generation.

Step F: mutation operation is carried out to the new population that intersection obtains.

Mutation operation whether operation is carried out for the individuality of first in population, generates the random number r of 0 to 1 ₂if, mutation probability p _m≤ r ₅, then a certain gene location P in Stochastic choice individuality _{bess, i, t} ^jcarry out mutation operation, P _{bess, i, t} ^jrepresent the energy storage charge-discharge electric power of i-th individual t in jth generation, the moment, t was for determine at random; If mutation probability p _m> r ₅, then mutation operation is not carried out.All aforesaid operations is carried out to all pairing individualities, to complete whole mutation process.

The method that in described Stochastic choice individuality, a certain gene location carries out mutation operation comprises the steps:

Determine a certain gene location in individuality by following formula, namely determine to carry out mutation operation to the accumulator system charge-discharge electric power of a certain moment t:

t＝[T*r ₆]

R ₆for the random number of 0 to 1 of generation.

During mutation operation, by following formula to P _{bess, i, t} ^jagain value:

P _bess，i，t ^j＝P _bess，chmax+r ₇×(P _{bess，dismax}-P _bess，chmax)

All aforesaid operations is carried out to all particles, to complete whole mutation process.

In formula, hop count (when optimizing hop count) when T is simulation, r ₇be the random number of 0 to 1, square bracket [] represent gets the most contiguous integer towards positive infinity; P _{bess, chmax}for the maximum permission charge power of accumulator system, be negative value, P _{bess, dismax}for the maximum permission discharge power of accumulator system, be on the occasion of.

Step G: the power due to battery energy storage system chooses fixing charge-discharge electric power value in Table 1, therefore there will not be the problem that power is out-of-limit, therefore it is out-of-limit only need to adjust state-of-charge.

After power is adjusted, through type (7) calculates the state-of-charge of battery energy storage system, and the state-of-charge constraint condition of through type (6) makes battery energy storage system whether exceed state-of-charge: if the state-of-charge of battery energy storage system should not in the scope that limits of formula, then adjust accordingly, if SOC _t> SOC _max, then P _{bess, t}=(SOC _t-1-SOC _max) C _bess/ Δ t, and as the P calculated _{bess, t}when not belonging to any number of table 1, P _{bess, t}get in table 1 from the numerical value calculated recently and than calculating the little state value of numerical value; SOC _t< SOC _min, P _{bess, t}=(SOC _t-1-SOC _min) C _bess/ Δ t, and as the P calculated _{bess, t}when not belonging to any number of table 1, P _{bess, t}get in table 1 from the numerical value calculated recently and than calculating the large state value of numerical value.

Step H: judge whether to reach iterations requirement or convergent requirement.

Iterations requires that this example can be set to 500 times in order to arrange according to actual requirement;

Convergent requirement is: max [f (P _bess ^j)]-min [f (P _bess ^j)] < α

In formula, P _bess ^jrepresent that jth is for all groups of individuals of population after j iteration, P _bess ^j={ P _{bess, 1} ^j, P _{bess, 2} ^j..., P _{bess, t} ^j..., P _{bess, T} ^j; F (P _bess ^j) represent the adaptive value of jth for all individualities, max [f (P _bess ^j)], min [f (P _bess ^j)] represent maximum, the minimum value of jth for adaptive values all in population, therefore namely this formula represents that jth is less than a certain value α for the difference of the maxima and minima of adaptive values all in population, α is a minimum number being greater than zero, and the minimum number of such as this example can adopt 0.01.

If meet one of them in two conditions, then stop iteration, export population at individual, if all do not meet, then forward step C to, continue iteration, until meet the demands.

Step I: in the end iteration obtains selecting optimum individuality in population, exports the energy storage charge-discharge electric power in each moment.

In the new population exported after iteration stopping the last time, find out the maximal value in all adaptive values, individuality (being optimum individual) corresponding to this maximal value, the target function value (being global optimum) of this optimum individual is calculated by above-mentioned formula (1), the wind power output climbing rate of optimum individual is calculated by above-mentioned formula (4), and the final charge-discharge electric power exporting battery energy storage system each moment.

Example explanation

The installed capacity of wind-driven power that the system of this example adopts is 100MW, and energy storage installed capacity is 10MW.

Choosing and setting of calculating parameter: rate for incorporation into the power network can be set to 0.5 yuan/kWh.

Energy storage power bracket; The charge-discharge electric power scope of energy storage is relevant with energy storage type.Common energy storage chemical cell, is more or less the same as the maximum charge performance number of lithium-ion energy storage battery and maximum discharge power value, and sodium-sulphur battery (NaS), it is little compared with maximum discharge power that maximum charge power generally sets; In addition, some novel energy-storing systems, as flywheel energy storage, superconduction capacitance energy storage etc., maximum charge-discharge electric power is substantially close.Accumulator system is not herein for a certain particular type, and energy storage discharge and recharge peak power is all set as 10MW.

In actual applications, energy storage controls to be generally that setting discharge and recharge working value is fixed, and selects for during operation.Herein by charge and discharge action restrictive condition discretize and number definition is corresponding state, specifically in table 1.Here each state in each stage all has 16 decision-makings, and control strategy is optimizes period all possible discharge and recharge combination.

The sequence table of table 1 battery energy storage system charge-discharge electric power

Domestic time-of-use tariffs are formulated according to the factor such as cost of electricity-generating, the level of economic development of various places.Due to actual energy storage operation cost calculation of complex, and relevant to energy storage type, consider herein only for illustration of the validity of this control strategy, thus conform to the principle of simplicity to consider to this problem, be set to 0.6 yuan/kWh in this example.In this example, the out-of-limit penalty term coefficient for wind power output climbing rate is set to 100 yuan/kWh.

Numerical results is shown in Fig. 2, Fig. 3 and table 2.

Fig. 2 is for optimizing front and back output of wind electric field comparison diagram, and count when in figure, horizontal ordinate represents, every 10 seconds time points, unit is min, and ordinate represents power, and unit is kW.As can be seen from Figure 2, after optimizing, the fluctuation of power curve is less.

Fig. 3 is that wind power output climbing rate optimizes forward and backward comparison diagram, counts when in figure, horizontal ordinate represents, every 10 seconds time points, and unit is min, and ordinate represents power, and unit is kW.Contrast known especially by Fig. 3, wind power output climbing rate after optimization is out-of-limit (namely obviously reducing more than the part of 6MW), associative list 2, wind power output climbing rate before optimization more than 6MW time count and reach 44, the ratio of counting time shared all reaches 36.67%, and after optimizing more than 6MW time count and only have 1 (and only having exceeded 0.4MW), the ratio of counting when putting all only have 0.83%, effect of optimization is fairly obvious.

Table 2 climbing rate optimizes front and back contrast table

The performance analysis of front and back optimized by table 3

Table 3 is for optimizing the performance analysis tables of data of front and back.Due to net proceeds=online income-discharge and recharge cost-penalty term, discharge and recharge cost=charging cost+electric discharge cost, number percent represents discharge and recharge cost, abandon eolian, penalty term and the every number percent with online income of net proceeds.Because the just wind power output state of 20 minutes of simulation, so charge value and the amount of money are all less.As can be seen from Table 3, although the online income before optimizing is larger, due to out-of-limit comparatively large, penalty term is also larger.Therefore, although paid certain energy storage discharge and recharge expense after optimizing, final net proceeds is higher before still comparatively optimizing.As can be seen from table 2 and table 3, the control method proposed in application the present invention, not only the control overflow of satisfied climbing rate constraint of exerting oneself is combined in wind storage, climbing rate wind-powered electricity generation being combined exert oneself controls within required scope, and take into account rate for incorporation into the power network, energy storage discharge and recharge cost, wind power output climbing rate out-of-limit time the factor such as rejection penalty, based on genetic optimization optimizing algorithm, achieve the maximum wind storing cogeneration systematic economy optimal control of wind storing cogeneration net proceeds and scheduling object, demonstrate feasibility and the validity of the method and control system thereof.

In sum, illustrate that method of the present invention is both realizing applying the control ability that accumulator system improves wind-powered electricity generation grade climbing performance, can ensure that again wind storage combines to exert oneself and meet the grid-connected requirement of stability bandwidth constraint, achieve the accumulator system multi objective control object considering accumulator system state-of-charge, the constraint of wind-powered electricity generation climbing rate, rate for incorporation into the power network, storage energy operation cost, wind storing cogeneration system grid connection maximizing generation profit etc.

Finally should be noted that: above embodiment is only in order to illustrate that technical scheme of the present invention is not intended to limit; in conjunction with above-described embodiment to invention has been detailed description; those of ordinary skill in the field are to be understood that: those skilled in the art still can modify to the specific embodiment of the present invention or equivalent replacement, but these amendments or change are all being applied among the claims awaited the reply.

Claims (18)

1. consider that the wind of rate for incorporation into the power network and genetic algorithm optimization stores up an electricity-generating control method, it is characterized in that, the method comprises the steps:

A, the parameter that genetic algorithm is set and rate for incorporation into the power network;

All individualities of B, generation initial population;

C, calculate the target function value of each individuality, and determine the adaptive value of each individuality further;

D, calculate the cumulative probability of each individual fitness, selection replicate run is carried out to each individuality, generates new population;

E, to the random pair between two of the individuality in new population, and interlace operation is carried out to often pair of individuality of successful matching;

F, according to mutation probability, carry out mutation operation to carrying out each individuality after interlace operation;

G, according to the constraint of the power bound of accumulator system and the constraint of state-of-charge bound, all individualities to be adjusted;

H, judge whether new population has met iterations requirement or convergent requirement, one of require as met these two, then stop iteration, export new population; Otherwise, jump to step C and continue iteration, till meeting one of these two requirements;

I, in new population, select optimum individual and global optimum, and export the accumulator system charge-discharge electric power in each moment.

2. the method for claim 1, is characterized in that, in steps A, the parameter of described genetic algorithm comprises crossover probability, mutation probability, iterations and Population Size.

3. the method for claim 1, is characterized in that, in step B, according to the first generation status switch of the accumulator system charge-discharge electric power of following formula in stochastic generation initial population, all N number of bodies, namely have individuality in population, for i-th individuality in 1st generation population representation as follows:

$P_{\mathrm{bess},i}^1=[{P_{\mathrm{bess},i,1}}^1,{P_{\mathrm{bess},i,2}}^1,...,{P_{\mathrm{bess},i,t}}^1,...,{P_{\mathrm{bess},i,T}}^1]$

In formula, i=1,2 ..., N, N represent Individual Size in population; [] represents the accumulator system vector that in hop count T, charge-discharge electric power forms when optimizing; P _{bess, i, t} ¹represent the charge-discharge electric power of i-th individual t in initial population; When charge-discharge electric power is that negative value represents that accumulator system is in charged state, on the occasion of time represent that accumulator system is in discharge condition, be that 0 expression battery energy storage system is in floating charge state, T for optimize time hop count;

According to the maximal value P of accumulator system charge and discharge power _{bess, chmax}, P _{bess, dismax}, stochastic generation is as shown in the formula each element P in all individualities _{bess, i, t} ¹:

P _bess，i，t ¹＝P _bess，chmax+r ₁×(P _{bess，dismax}-P _bess，chmax)

In formula, P _{bess, chmax}for negative value, represent the maximum permission charge power of accumulator system, P _{bess, dismax}for on the occasion of, represent accumulator system maximum permission discharge power; r ₁represent the random number between 0 to 1.

4. the method for claim 1, is characterized in that, in described step C,

For i-th individual adaptive value f _i ^jbe calculated as follows:

${f_i}^j=F_i^j-F_{\min }^j$

In formula, represent that jth is for i-th individuality in population target function value, represent the minimum value of jth for individual goal functional values all in population, described target function value asked for by following formula:

$\max F_i^j=\underset{t=1}{\overset{T}{\mathrm{\&Sigma;}}}[c_{\mathrm{grid}}\&CenterDot;P_{\mathrm{grid},i,t}^j\&CenterDot;\mathrm{\&Delta;t}-c_{\mathrm{bess}}\&CenterDot;|P_{\mathrm{bess},i,t}^j|\&CenterDot;\mathrm{\&Delta;t}-M\&CenterDot;P_{\mathrm{punish},i,t}^j\&CenterDot;\mathrm{\&Delta;t}]$

In formula, hop count when T is for optimizing; c _gridfor wind-powered electricity generation rate for incorporation into the power network; for according to the charge-discharge electric power P of jth for accumulator system t in population i-th individuality _{bess, i, t} ^jthe wind storage commingled system calculated goes out force value to electrical network; Δ t is the interval that wind power output is measured; c _bessfor the unit operation cost of accumulator system; P _{bess, i, t} ^jfor jth is for the charge-discharge electric power of accumulator system t in population i-th individuality; P _{bess, i, t} ^jfor negative value, represent that accumulator system is in charged state; P _{bess, i, t} ^jfor on the occasion of, represent accumulator system be in discharge condition; P _{bess, i, t} ^jbe 0, represent that accumulator system is in floating charge state; M is the out-of-limit penalty term coefficient of wind power output climbing rate; for according to the charge-discharge electric power P of jth for accumulator system t in population i-th individuality _{bess, i, t} ^jpenalty term when the wind power output climbing rate calculated is out-of-limit.

5. method as claimed in claim 4, is characterized in that,

Described wind storage commingled system goes out force value to electrical network for:

$P_{\mathrm{grid},i,t}^j=P_{\mathrm{wind},t}+{P_{\mathrm{bess},i,t}}^j$

Described wind power output climbing rate for:

$P_{\mathrm{ramp},i,t}^j=\left\{\begin{array}{cc}\left|{\mathrm{\&Delta;P}}_{\max +,i,t}^j\right|,& \mathrm{if}\left|{\mathrm{\&Delta;P}}_{\max +,i,t}^j\right|\&GreaterEqual;\left|{\mathrm{\&Delta;P}}_{\max -,i,t}^j\right|\\ \left|{\mathrm{\&Delta;P}}_{\max -,i,t}^j\right|,& \mathrm{if}\left|{\mathrm{\&Delta;P}}_{\max +,i,t}^j\right|<\left|{\mathrm{\&Delta;P}}_{\max -,i,t}^j\right|\end{array}\right.$

Penalty term when described wind power output climbing rate is out-of-limit for:

$P_{\mathrm{punish},i,t}^j=\left\{\begin{array}{cc}{P}_{\mathrm{ramp},i,t}^j-P_{\mathrm{stad}},& \mathrm{if}{P}_{\mathrm{ramp},i,t}^j-P_{\mathrm{stad}}>0\\ 0,& \mathrm{if}{P}_{\mathrm{ramp},i,t}^j-P_{\mathrm{stad}}\&le;0\end{array}\right.$

In formula, P _{wind, t}for the wind power output value of t; for the wind power output climbing rate calculated for the charge-discharge electric power of accumulator system t in population i-th individuality according to jth; for going out the absolute value of the difference of force value for interior maximum wind of wind power output value and [t, t-K] interval that the charge-discharge electric power of accumulator system t in population i-th individuality calculates according to jth; for the wind power output value calculated for the charge-discharge electric power of accumulator system t in population i-th individuality according to jth interval with [t, t-K] in the absolute value of difference of minimum wind power output value; K is the time period of set investigation when suppressing wind power output climbing rate; P _stadfor wind power output climbing rate standard value.

6. the method for claim 1, is characterized in that, in step D, the cumulative probability of described individual fitness is:

$p_i=\left\{\begin{array}{cc}{f_i}^j/{\mathrm{\&Sigma;}}_{i=1}^N{f_i}^j,& \mathrm{ifi}=1\\ p_{(i-1)}+{f_i}^j/{\mathrm{\&Sigma;}}_{i=1}^N{f_i}^j,& \mathrm{ifi}\&GreaterEqual;2\end{array}\right.$

In formula, p _i, p _(i-1)be i-th, an i-1 individual cumulative probability; f _i ^jbe i-th individual adaptive value;

for the adaptive value sum of individualities all in initial population; N is sum individual in initial population;

According to following formula, selection replicate run is carried out to all individualities, generates new population:

$P_{\mathrm{bess},i}^j=\left\{\begin{array}{cc}{P}_{\mathrm{bess},1}^{j-1},& \mathrm{if}{r}_2\&le;p_1\\ P_{\mathrm{bess},k}^{j-1},& \mathrm{if}{p}_k<r_2\&le;p_{k+1}\end{array}\right.$

In formula, represent that jth is for the individuality of i-th in population, represent that j-1 is for the 1st in population, a k individuality respectively, r ₂it is the random number between 0 to 1.

7. the method for claim 1, is characterized in that, in step e, by comparing crossover probability p _cwith random number r ₃size, determine whether that every a pair individuality after to random pair carries out interlace operation, described r ₃for the random number of 0 to 1 of stochastic generation;

If crossover probability p _c≤ r ₃, then determine to carry out interlace operation to a certain gene location in individuality by following formula, namely determine to carry out interlace operation to the accumulator system charge-discharge electric power of a certain moment t:

t＝[T*r ₄]

In formula, hop count when T is for optimizing; r ₄for the random number of 0 to 1 of generation.

If crossover probability p _c> r ₃, then interlace operation is not carried out.

8. the method for claim 1, is characterized in that, in step F, by comparing mutation probability p _mwith random number r ₅size, determine whether carrying out mutation operation, described r to carrying out each individuality after interlace operation in new population ₅for the random number of stochastic generation 0 to 1;

If mutation probability p _m≤ r ₅, then determine to carry out mutation operation to gene location a certain in individuality by following formula, namely determine to carry out mutation operation to the accumulator system charge-discharge electric power of a certain moment t:

t＝[T*r ₆]

In formula, hop count when T is for optimizing; r ₆for the random number of 0 to 1 of generation.

If mutation probability p _m> r ₅, then mutation operation is not carried out.

9. method as claimed in claim 8, is characterized in that, during mutation operation, by following formula to all particles again value, to complete whole mutation process:

P _bess，i，t ^j＝P _bess，chmax+r ₇×(P _{bess，dismax}-P _bess，chmax)

In formula, hop count when T is simulation; r ₇it is the random number of 0 to 1; P _{bess, chmax}for the maximum permission charge power of accumulator system, be negative value; P _{bess, dismax}for the maximum permission discharge power of accumulator system, be on the occasion of.

Power, be negative value; P _{bess, dismax}for the maximum permission discharge power of accumulator system, be on the occasion of.

10. the method for claim 1, is characterized in that, in step G, comprises the method that all individualities adjust:

G1) each moment accumulator system power upper limit value and lower limit value P is read _minand P _max

And by the power bound constraint condition of following formula to the charge-discharge electric power P of current accumulator system _{bes, t}retrain:

P _min≤P _bess，t≤P _max

If P _{bess, t}do not exceed above-mentioned constraint condition scope, then do not change P _{bess, t}, and jump to step G2; If P _{bess, t}> P _max, make P _{bess, t}=P _max; If P _{bess, t}< P _min, make P _{bess, t}=P _min;

G2) state-of-charge of accumulator system is calculated by following formula:

${\mathrm{SOC}}_t={\mathrm{SOC}}_{t-1}-\frac{P_{\mathrm{bess},t}*\mathrm{\&Delta;t}}{C_{\mathrm{bess}}}\&times;100\%$

And retrained by the state-of-charge of state-of-charge bound constraint condition to current accumulator system of following formula:

SOC _min≤SOC _t≤SOC _max

If SOC _t> SOC _max,make P _{bess, t}=(SOC _t-1-SOC _max) C _bess/ Δ t, if P _{bess, t}actual computation value when being not equal to arbitrary predetermined power value, then get in predetermined power value with actual computation value closest to and be worth little value than actual computation;

If SOC _t< SOC _min, make P _{bess, t}=(SOC _t-1-SOC _min) C _bess/ Δ t, if P _{bess, t}actual computation value when being not equal to arbitrary predetermined power value, then get in predetermined power value with actual computation value closest to and the value larger than actual computation value.

In formula, P _minfor negative value, the lower limit representing charge power, P _maxfor on the occasion of, represent the upper limit of discharge power; SOC _t, SOC _t-1for the state-of-charge of t, t-1 moment accumulator system; SOC _min, SOC _maxbe respectively state-of-charge lower limit, the higher limit of accumulator system; P _{bess, t}for t accumulator system charge-discharge electric power; Δ t is the interval that wind power output is measured.C _bessfor the rated capacity of accumulator system.

11. the method for claim 1, is characterized in that, in step H,

Described iterations requires to arrange according to actual conditions;

Described convergent requirement is:

max[f(P _bess ^j)]-min[f(P _bess ^j)]＜α

P _bess ^j＝{P _bess，1 ^j，P _bess，2 ^j，…，P _bess，t ^j，…，P _bess，T ^j}

In formula, P _bess ^jrepresent that jth is for all groups of individuals of population after j iteration; F (P _bess ^j) represent the adaptive value of jth for all individualities; Max [f (P _bess ^j)], min [f (P _bess ^j)] represent maximum, the minimum value of all adaptive values in jth generation; α is a minimum number being greater than zero.

12. the method for claim 1, is characterized in that, in step I, in the new population exported after iteration the last time, find out the maximal value in all adaptive values, the individuality corresponding to this maximal value is optimum individual, and the target function value of described optimum individual is global optimum.

Consider that the wind of rate for incorporation into the power network and genetic algorithm optimization stores up power-generating control system for 13. 1 kinds, it is characterized in that, this system comprises:

Population generation module, for arranging the parameter of genetic algorithm, and generates initial population;

Computing module, for calculating adaptive value and the cumulative probability thereof of each individuality in initial population, carries out selection replicate run to each individuality, generates new population;

Cross module, for the random pair between two of the individuality in new population, and carries out interlace operation to the individuality of often pair of successful matching;

Variation module, for carrying out mutation operation to carrying out each individuality after interlace operation; With

Execution module, for judging whether new population meets iterations requirement or convergent requirement, until export new population after meeting arbitrary requirement, and find out the corresponding individual position of global optimum wherein, the position of described individuality is the charge-discharge electric power of the corresponding accumulator system of this individuality.

14. the system as claimed in claim 1, is characterized in that, described population generation module comprises:

Optimum configurations submodule, for arranging the parameter of genetic algorithm according to actual conditions;

First calculating sub module, for according to accumulator system charge-discharge electric power state, all individualities in stochastic generation initial population.

15. the system as claimed in claim 1, is characterized in that, described computing module comprises:

Second calculating sub module, for calculate wind storage commingled system out-of-limit to the wind power output value of electrical network, wind power output climbing rate and wind power output climbing rate time penalty term, and determine the adaptive value of each individuality further by the objective function F of wind storage commingled system generating optimization;

Cumulative probability calculating sub module, for asking for the cumulative probability of each individual fitness according to adaptive value;

Replicon module, for the comparative result according to cumulative probability and random number, adjusts each individuality and generates new population.

16. the system as claimed in claim 1, is characterized in that, described Cross module comprises:

First comparison sub-module, for by comparing crossover probability p _cwith random number r ₃size, determine whether carrying out interlace operation to often pair of individuality of successful matching in population, described random number is the random number of 0 to 1 of stochastic generation;

Interlace operation submodule, if desired carries out interlace operation, then first determine to carry out interlace operation to the accumulator system charge-discharge electric power in which moment; And when interlace operation, the power in the determined moment of the individuality of successful matching is exchanged, to complete interlace operation.

17. the system as claimed in claim 1, is characterized in that, described variation module comprises:

Second comparison sub-module, by comparing mutation probability p _mwith random number r ₅size, determine whether carrying out mutation operation, described r to individualities all in population ₅for the random number of 0 to 1 of stochastic generation;

Mutation operation submodule, if desired carries out mutation operation, then first determine to carry out mutation operation to the accumulator system charge-discharge electric power in which moment; And when mutation operation, carry out the individuality after interlace operation carry out value again, to complete mutation operation to all.

18. the system as claimed in claim 1, is characterized in that, described execution module comprises:

Judge module, for judging whether new population meets iterations requirement or convergent requirement, described iterations requires to arrange according to actual conditions, and described convergent requirement is that all ideal adaptations are worth the difference of maximal and minmal value to be less than a minimum number being greater than zero;

Result output module, for selecting global optimum individuality from the new population meeting iterations requirement or convergent requirement, and recalculates the adaptive value of this individuality by computing module, and exports the accumulator system charge-discharge electric power corresponding to this individuality.