CN110518634A - Wind-powered electricity generation field control method is accessed based on the batteries to store energy system for improving exponential smoothing - Google Patents

Abstract

The invention discloses a method for controlling the connection of a battery energy storage system to a wind farm based on an improved exponential smoothing method, comprising the following steps: S1. Obtaining historical wind power data, and using the improved exponential smoothing method to predict the output power in the next period; S2. Calculating wind power output fluctuations Whether the rate meets the industry standard, if not, go to step S3; if so, end; S3, consider the battery state of charge and the standard value of the volatility rate as constraints, control the battery energy storage system to charge and discharge. The present invention introduces an adaptive particle swarm algorithm in the process of exponentially smoothing wind power fluctuations, realizes the adaptive selection of smoothing coefficients, and improves the effect of traditional quadratic exponential smoothing. The simulation results show that in a wind farm with a certain energy storage capacity, the control strategy of the present invention has a better smoothing effect.

Description

Control method of battery energy storage system connected to wind farm based on improved exponential smoothing method

technical field

The invention relates to the technical field of wind farm control, in particular to a method for controlling the connection of a battery energy storage system to a wind farm based on an improved exponential smoothing method.

Background technique

Wind power is an important part of my country's energy strategy. It has developed rapidly and provided a large amount of clean energy for social development. Because wind power is easily affected by factors such as climate and environment, its output has great randomness and volatility, and there is a certain error in wind power prediction at present. The quality of power supply and safe operation of the system have a great impact, which seriously restricts the large-scale application of wind power generation. How to effectively stabilize the output power of wind farms is of great significance. Equipping an energy storage unit in a wind farm can effectively reduce the fluctuation of the output power of the wind farm, convert wind power into dispatchable power to a certain extent, and help reduce the impact of wind power on the power system. The battery has the characteristics of simple maintenance, long service life, stable quality and high reliability. With the development of battery technology and power electronics technology, battery energy storage systems have been widely used in smoothing wind power fluctuations.

Among them, exponential smoothing method is an important method for time series analysis, and has a wide range of applications in economic forecasting, power metering equipment demand forecasting, and energy conservation. The smoothing coefficient is one of the important parameters in the exponential smoothing algorithm, and its selection determines the accuracy of the prediction. In the traditional method, the smoothing coefficient is manually specified, and its disadvantage is that it relies on artificial experience and is not dynamic. Commonly used automatic optimization algorithms for smooth coefficients include: golden section method, Fibonacci method, tangent method, and dichotomy method. These automatic optimization algorithms can obtain the optimal value of the smoothing coefficient in a short time, but only when the smoothing coefficient is a single extreme value. When the smoothing coefficient is a multi-extremal value, these algorithms cannot guarantee The value converges to the optimal smoothing coefficient.

SUMMARY OF THE INVENTION

The purpose of the present invention is to provide a control method based on an improved exponential smoothing method for battery energy storage systems to be connected to wind farms, and to construct a wind power-battery energy storage system (BESS) joint operation system. The exponential smoothing method controls the energy storage system, and the particle swarm optimization algorithm is introduced into the process of solving the smoothing coefficient, and the reasonable selection of the smoothing coefficient is realized with the minimum sum of squares of prediction errors (SSE) as the constraint condition. The national standard for grid volatility is constrained to achieve reasonable access to wind power.

To achieve the above object, the present invention adopts the following technical solutions:

The control method of battery energy storage system connected to wind farm based on improved exponential smoothing method includes the following steps:

S1. Obtain the historical data of wind power, and use the improved exponential smoothing method to predict the output power in the next period;

S2. Calculate whether the wind power output volatility complies with the industry standard, if not, go to step S3; if it complies, end;

S3. Considering the battery state of charge and the standard value of the fluctuation rate as constraints, control the battery energy storage system to charge and discharge.

Further, the acquired wind power historical data and the improved exponential smoothing method are used to predict the output power in the next period, specifically including:

Obtain all historical data of wind power, and obtain a smooth prediction result through the weighted average method;

By using the minimum sum of squares of prediction errors as the evaluation index, the adaptive particle swarm algorithm is used to solve the smoothing coefficient of the quadratic smoothing model, and the output power sequence prediction of the next period is completed.

Further, the solution of the smoothing coefficient of the quadratic smoothing model is realized by the adaptive particle swarm algorithm, and the output power sequence prediction of the next period is completed, which specifically includes:

Input the original sampling data of the wind farm and initialize the particle swarm;

Calculate and evaluate the fitness value of each particle, and solve the optimal solution p _zy of the group and the optimal solution p _dq of the current position of the particle;

Update particle information based on velocity, position and weight factor formulas;

The calculation of mutation probability is completed according to the calculation result of the variance of the adaptive degree;

Judging the number of iterations, complete the iterative output group optimal solution p _zy ; otherwise, continue the iteration.

Further, the calculation of whether the wind power output volatility complies with industry standards specifically includes:

Calculate whether the capacity of the battery energy storage system equipped in the wind farm, the 1-min power fluctuation limit of grid-connected wind power, and the 10-min power fluctuation limit of grid-connected wind power meet the following table:

.

Further, considering the battery state of charge and the standard value of the fluctuation rate as constraints, the battery energy storage system is controlled to charge and discharge, specifically including:

The control system judges that the 1min or 10min power fluctuation of the grid-connected wind power exceeds the specified value, and the battery energy storage system charges and discharges to smooth the difference between the target and the actual wind power output power. The battery energy storage system needs to absorb or release the power and the target of the system. The output power can be respectively recorded as:

When P _{out_W,k} >P _{out_S,k} , the battery energy storage system absorbs power; when P _{out_W,k} <P _{out_S,k} , the battery energy storage system releases power, and its value is the difference between the smoothing output and the wind power. difference.

Further, the battery energy storage system needs to calculate the power absorbed or released and the target output power of the system, including:

Input the power data of the wind farm for a certain period of time, calculate the power fluctuation rate and the output or absorbed power value of the battery energy storage system required to stabilize the fluctuation;

Consider the output power fluctuation of the wind farm after the output power of the battery energy storage system, and judge whether the fluctuation rate at this time meets the requirements of the industry standard table. If not, adjust the charging and discharging power of the battery energy storage system. the state of charge of the battery;

Determine whether the state of charge of the battery meets the requirements, if not, end directly; if so, proceed to the next iteration until the end condition is met, and output the output power of the wind farm and the capacity of the corresponding battery energy storage system that needs to be charged or discharged after the iteration. .

The effect provided in the summary of the invention is only the effect of the embodiment, rather than all the full effects of the invention, a technical scheme in the above-mentioned technical solutions has the following advantages or beneficial effects:

The present invention introduces an adaptive particle swarm algorithm in the process of exponentially smoothing the fluctuation of wind power, aiming at the problem that the "particles" in the traditional particle swarm algorithm may "oscillate" near the global optimal solution in the control strategy. The inertia weight is adjusted in an adaptive way to improve the convergence of the algorithm. The adaptive selection of smoothing coefficient is realized, and the effect of traditional quadratic exponential smoothing is improved. The simulation results show that, in the wind farm with the determined energy storage capacity, the control strategy of the present invention has a better smoothing effect.

Description of drawings

1 is a flow chart of a method according to an embodiment of the present invention;

Fig. 2 is the principle diagram of the wind farm operation system based on the battery energy storage system of the present invention;

3 is a schematic diagram of the control strategy of the battery energy storage system of the present invention;

Fig. 4 is the wind power fluctuation control flow diagram of the present invention;

Fig. 5 is the original power fluctuation curve of the wind farm;

Fig. 6 is a traditional quadratic smoothing method to suppress the effect diagram;

Fig. 7 is the smoothing effect diagram of the improved quadratic smoothing method based on the adaptive particle swarm algorithm;

Figure 8 is a 1min volatility curve of the basic quadratic smoothing method;

Figure 9 is a graph of the 1min volatility curve of the improved quadratic smoothing method;

Figure 10 is a 10-min volatility curve of the traditional quadratic smoothing method;

Figure 11 is a 10-min volatility curve of the improved quadratic smoothing method.

Detailed ways

In order to clearly illustrate the technical features of this solution, the present invention will be described in detail below through specific embodiments and in conjunction with the accompanying drawings. The following disclosure provides many different embodiments or examples for implementing different structures of the invention. In order to simplify the disclosure of the present invention, the components and arrangements of specific examples are described below. Furthermore, the present invention may repeat reference numerals and/or letters in different instances. This repetition is for the purpose of simplicity and clarity and does not in itself indicate a relationship between the various embodiments and/or arrangements discussed. It should be noted that the components illustrated in the figures are not necessarily drawn to scale. Descriptions of well-known components and processing techniques and processes are omitted from the present invention to avoid unnecessarily limiting the present invention.

As shown in Figure 1, the method for controlling the connection of battery energy storage systems to wind farms based on the improved exponential smoothing method includes the following steps:

S1. Obtain the historical data of wind power, and use the improved exponential smoothing method to predict the output power in the next period;

S2. Calculate whether the wind power output volatility complies with the industry standard, if not, go to step S3; if it complies, end;

S3. Considering the battery state of charge and the standard value of the fluctuation rate as constraints, control the battery energy storage system to charge and discharge.

As shown in Figure 2, the energy storage battery is reasonably configured in the wind farm, and the control strategy is formulated to coordinate the output with the wind farm to form a wind power-battery energy storage system joint operation system, and P _{out_B} is the battery energy storage system. Charge and discharge power, P _{out_W} is the output power of the wind farm, and P _{out_S} is the combined output power of the wind power-battery energy storage system combined operation system. According to the law of conservation of energy:

P _{out_T} =P _{out_W} +P _{out_B} (1)

The battery energy storage system can be charged and discharged as required. The present invention stipulates that P _{out_B} > 0 means that the battery energy storage system is discharged. If P _{out_B} < 0, it means that the battery energy storage system is charged. In a certain period of time, the battery energy storage system is charged. The calculation relationship between the discharge power output limit and SOC is:

In the formula: △t is the sampling time interval, and the values are 10min and 1min in the present invention; P _{cmaxb_B} , P _{dmaxb_B} and P _{clim_B(n)} , P _{dlim_B(n)} respectively represent the extreme value of the charging and discharging power of the energy storage system and a certain Limits of charging and discharging power within a period; P _{n_B(n)} and _{En_B} are the rated values of system power and capacity, respectively; S _{max_B} and S _{min_B} are the upper and lower limits of system SOC; S _B(n-1) and S _{B (n)} is the SOC value of the current and last time state; σ _B is the self-discharge power; η _{c_B} and η _{d_B} are the efficiencies of the energy storage system charging and discharging process, respectively.

By configuring the battery energy storage system in the wind farm and formulating a coordinated control strategy, the power fluctuation of the output power P _{out_T} of the entire system is limited within a reasonable range to meet the grid-connection requirements of the wind farm, and it is a large-scale wind farm. Grid connection creates conditions.

In step S1, the historical data of wind power is obtained, and the improved exponential smoothing method is used to predict the output power of the next period, which specifically includes:

Obtain all historical data of wind power, and obtain a smooth prediction result through the weighted average method;

By using the minimum sum of squares of prediction errors as the evaluation index, the adaptive particle swarm algorithm is used to solve the smoothing coefficient of the quadratic smoothing model, and the output power sequence prediction of the next period is completed.

The exponential averaging method is based on all the historical data of the forecast object, and the forecast result is obtained by means of weighted average. The recursive formula is:

In the formula: α is the smoothing coefficient, n is the smoothing index, x _t+1 is the predicted value, and x _t , x _t-1 ...x _tn is the observed value.

From the above formula: In the process of solving the predicted value, the size of α can determine the proportion of each observation data in the prediction process, that is, the smoothing coefficient will have a greater impact on the smoothing result. In this embodiment, on the basis of one-time smoothing, the second-order exponential smoothing prediction is realized by adjusting the predicted value. The model for quadratic exponential smoothing is as follows:

basic formula

prediction formula

where the prediction parameter

It can be seen from the above formula that the calculation of related parameters can be completed by smoothing the output x _t ⁽¹⁾ and x _t ⁽²⁾ , where Indicates that the t+Tth value is predicted. From the quadratic smoothing model, the factors that affect the predicted value are:

①Smoothing coefficient α, its value directly reflects the changing trend of data in different time periods.

②Initial value x ₀ . Engineering practice shows that there are many time series of observation values, and the influence of initial value on prediction results can be ignored. The present invention conforms to this feature, that is, the influence of initial value x ₀ on subsequent prediction results is ignored.

Based on the above-mentioned prediction model, the present invention adopts the minimum sum of squares of prediction errors (SSE) as the evaluation index, realizes the solution of the smoothing coefficient with the adaptive particle swarm algorithm, and completes the secondary smoothing of the entire event sequence.

Particle swarm optimization (PSO) is a stochastic evolution calculation method developed on the basis of bird predation research. The algorithm has the characteristics of easy implementation, fast convergence speed and high precision, and has been widely used in engineering practice.

In each iteration process, each particle has to complete the update of the position and velocity, and the update rule is as follows:

In the formula: p _dq ^k is the optimal position of the current particle; p _zy ^k is the historical optimal position of the particle swarm; c ₁ and c ₂ are learning factors; ω is the inertia weight; r1 and r2 are between 0 and 1 random numbers between.

In the calculation process, the inertia weight ω plays the role of keeping the particle moving, ensuring its tendency to expand the search space. In the application process, particles may "oscillate" near the global optimal solution. In order to solve this problem and improve the convergence of the algorithm, the present invention adopts an adaptive way to adjust the inertia weight, so that it can linearly adjust the inertia weight in the algorithm iteration process. reduce, that is:

ω=ω _max -n _c (ω _max -ω _min )/n _cmax (9)

In the formula: ω _min and ω _max are the extreme values of the inertia weight ω, respectively; n _c and n _cmax are the current iteration number and the maximum iteration number, respectively.

The particle swarm algorithm is easy to fall into the local optimal solution in the iterative calculation process, and the present invention realizes the mutation probability of the current optimal particle with the population fitness variance and the size of the current optimal solution in the process of the algorithm. Through this operation, the algorithm can realize the jumping out after falling into the local optimal solution and avoid the occurrence of premature problems.

In the process of the algorithm, the change trend of the fitness value of all particles in the group can be calculated to realize the quantitative analysis of the aggregation degree of each particle. The group fitness variance is:

In the formula: n is the total number of particles; f _i is the fitness of the ith particle; f _a is the average fitness of the current particle swarm; σ ² is the population fitness variance; f is used to limit the size of σ ² Calibration factor, its calculation formula is:

From the above formula: σ ² directly reflects the "convergence" degree of the entire particle swarm; the smaller the value is, the particle swarm tends to converge; otherwise, the entire particle swarm is in the stage of random convergence.

In order to avoid the search stagnation of the entire particle swarm, the present invention uses the size of the population fitness variance to determine the "mutation" mutation probability P _k of the particle swarm in a new direction. The formula ^[13] is as follows:

In the formula: P _k is the mutation probability that the k-th iteration process makes the global optimal; σ _k is the adaptive degree variance of the population in the k-th iteration; P _max and P _min are the extreme values of the mutation probability, respectively.

The present invention introduces random disturbance in the process of mutation operation of p ^k _zy , and its calculation formula is:

In the formula: η is a random variable, which conforms to the Guass(0,1) distribution.

Based on the adaptive particle swarm algorithm, the process of solving the smoothing coefficient is as follows:

1) Initialization, input the observation data, initialize the initial conditions of the algorithm, determine the overall size Q of the particle swarm, the maximum number of iterations N, the inertia weight ω and the learning factors c ₁ , c ₂ .

2) Generate random particles in the feasible region, complete the setting of the initial position x _i and the initial velocity vi of each particle _, and set the individual optimal solution and the global optimal solution of each initial particle to a sufficiently large value.

3) For the particles generated in 2), calculate the fitness value of each particle, determine the smallest value as the group optimal solution p _zy , and set p _dq as the optimal solution for the current position of the particle.

4) The counter is updated, the inertia weight is calculated according to formula (9), and the calculation of particle position x _i and velocity vi is completed according to formula (8 ₎ ; in the calculation process, the particles that cross the boundary can be processed according to the boundary mutation strategy, that is, if _vi > v _max , then v _i =v _max , if v _i <-v _max , then v _i =-v _max .

5) Re-evaluation of the particle fitness value, compare the optimal solution p _dq of the current individual with the fitness value f _s ( _xi ) of each particle, if f _s ( _xi )<p _dq , then p _dq =f _s ( _xi ), x _pi =xi; if f _smin <p _zy , that is, the optimal solution of the current generation is smaller than the optimal solution of the previous generation, then p _zy =f _smin .

6) Particle swarm mutation operation. According to the calculation result of the variance of the adaptive degree, the calculation of the mutation probability is completed according to the formula (12), and the random number r∈[0,1] is generated. If r<P _k , the mutation operation is completed according to the formula (13), otherwise, the to step 6).

7) Judging the number of iterations, if the iteration is completed, go to step 8), otherwise go to step 4).

8) Complete the output of the optimal solution p _zy .

Wind power generation is intermittent and fluctuating, and its grid connection will affect the stability of grid operation. In order to reduce the impact of wind power grid connection, its power fluctuations must be suppressed. According to the regulations on the connection of wind farms to the power system, the maximum power fluctuations are shown in the following table:

Table 1 National standards for the limit of power fluctuations in grid-connected wind farms

Tab.1 The national standard of active power in wind farm

In step S2, calculating whether the wind power output fluctuation rate complies with the industry standard, specifically including: calculating the capacity of the battery energy storage system equipped with the wind farm, the 1-min power fluctuation limit of the grid-connected wind power, and whether the 10-min power fluctuation limit of the grid-connected wind power meets the table. 1.

As shown in Figure 3, the battery energy storage system is equipped in the wind farm, and the battery energy storage system needs to be controlled. P _{out_W} is the output power of the wind farm at time t; P _{out_S} is the power output value after smooth calculation; P _{out_c} is the whole The difference between the output power of the system and the wind power output power; P _{out_T} is the output power of the entire system; δ is the wind power fluctuation rate.

When the control system detects that the 1min or 10min power fluctuation of the grid-connected wind power exceeds the specified value, the battery energy storage system will charge and discharge to smooth the difference between the target and the actual wind power output power, so as to achieve the purpose of suppressing the output power fluctuation of the wind farm. The power that the energy storage system needs to absorb or release and the target output power of the system can be respectively recorded as:

When P _{out_W,k} >P _{out_S,k} , the battery energy storage system absorbs power, and when P _{out_W,k} <P _{out_S,k} , the battery energy storage system releases power, and its value is equal to the difference between the output and the wind power value.

According to the wind farm grid-connected maximum power fluctuation and system smoothing requirements, the control process is shown in Figure 4: In Figure 4: _N is the maximum number of iterations, δ _TE is the power fluctuation of TE, and δ ^up _TE is the power within _TE time The maximum value of the fluctuation. The input of Figure 4 is the power data of the wind farm for a certain period of time, and the fluctuation rate of these powers is calculated. At the same time, the output or absorbed power value of the battery energy storage system is required to smooth these fluctuations. The next step is to consider the output power of the battery energy storage system. The output power of the wind farm fluctuates, and judge whether the fluctuation rate at this time meets the requirements of Table 1. If not, adjust the charging and discharging power of the battery energy storage system. If it meets the requirements, calculate the state of charge (SOC) of the battery at this time, and Execute the next step to judge the state of charge of the battery. If the requirements are not met, end directly. If the requirements are met, proceed to the next iteration until the end conditions are met. After the iteration, the output power of the wind farm and the corresponding capacity of the battery energy storage system to be charged or discharged can be obtained.

A specific calculation example is given below.

Through the data simulation analysis of the measured data of a large foreign wind farm, the rated grid-connected power is 35 MW, the capacity of the battery energy storage system is 10 MW, the sampling period of the system is 1 min, and the SOC value range of the battery energy storage system is [0.3, 1]. The initial power fluctuation curve is shown in Figure 5:

This example is based on the national wind power grid-connected standard, and the original grid-connected power of the wind farm is stabilized in two ways. The results are shown in Figures 6 and 7.

In order to better compare the smoothing effect, the smoothed 1min and 10min volatility are calculated, as shown in Figure 8 and Figure 9, respectively.

Table 2 Comparison of smoothing effects

Table 2 shows the comparison of the smoothing effect between the basic quadratic smoothing method and the improved quadratic smoothing method proposed by the present invention. It can be observed from Table 2 and Fig. 6 to Fig. 11 that in a wind farm with a certain energy storage capacity, through two Sub-smoothing can reduce the power fluctuation of wind farm grid connection, and 1min and 10min power fluctuation can meet the national standard of wind power grid connection. And the effect of the quadratic smoothing method based on the adaptive particle swarm algorithm proposed by the present invention is more ideal than the traditional quadratic smoothing method, which proves the correctness and effectiveness of the control strategy proposed by the present invention.

Although the specific embodiments of the present invention have been described above in conjunction with the accompanying drawings, they do not limit the scope of protection of the present invention. Those skilled in the art should understand that on the basis of the technical solutions of the present invention, those skilled in the art do not need to pay creative work. Various modifications or deformations that can be made are still within the protection scope of the present invention.

Claims (6)

1. accessing wind-powered electricity generation field control method based on the batteries to store energy system for improving exponential smoothing, characterized in that including following step It is rapid:

S1, wind-powered electricity generation historical data is obtained, predicts subsequent period output power using exponential smoothing is improved；

S2, it calculates whether wind-powered electricity generation output pulsation rate meets professional standard, if do not met, is transferred to step S3；If met, knot Beam；

S3, storage battery charge state and stability bandwidth standard value are considered for constraint condition, control energy-storage system of accumulator carries out charge and discharge Electricity.

2. wind-powered electricity generation field control method is accessed based on the batteries to store energy system for improving exponential smoothing as described in claim 1, It is characterized in, the acquisition wind-powered electricity generation historical data, predicts subsequent period output power using exponential smoothing is improved, specifically include:

Wind-powered electricity generation whole historical data is obtained, a smoothing prediction result is obtained by weighted average method；

It is evaluation index by using minimum Prediction sum squares, secondary smoothing model is realized with APSO algorithm Smoothing factor solves, and completes subsequent period output power sequence prediction.

3. wind-powered electricity generation field control method is accessed based on the batteries to store energy system for improving exponential smoothing as claimed in claim 2, It is characterized in, it is described to realize that the smoothing factor of secondary smoothing model solves with APSO algorithm, complete subsequent period output Power sequence prediction, specifically includes:

Wind power plant original sampling data is inputted, population is initialized；

The adaptive value of each particle of Calculation Estimation solves group optimal solution p_zyThe optimal solution p of current position with particle_dq；

Particle information is updated according to speed, position and weight factor formula；

The calculating of mutation probability is completed according to adaptive response variance calculated result；

The number of iterations judgement completes iteration and exports group optimal solution p_zy；Otherwise continue iteration.

4. wind-powered electricity generation field control method is accessed based on the batteries to store energy system for improving exponential smoothing as described in claim 1, It is characterized in, whether the calculating wind-powered electricity generation output pulsation rate meets professional standard, it specifically includes:

Calculate energy-storage system of accumulator capacity that wind power plant is equipped with, the 1min power swing limit value of grid connected wind power, grid connected wind power Whether 10min power swing limit value meets following table:

。

5. wind-powered electricity generation field control method is accessed based on the batteries to store energy system for improving exponential smoothing as claimed in claim 4, It is characterized in, the consideration storage battery charge state and stability bandwidth standard value are constraint condition, and control energy-storage system of accumulator carries out Charge and discharge specifically include:

Control system judges that 1min the or 10min power swing of grid connected wind power is more than specified value, then energy-storage system of accumulator carries out Charge and discharge to stabilize the difference of target He practical wind power output power, power that energy-storage system of accumulator needs to absorb or discharge and The target output of system can be denoted as respectively:

Work as P_{Out_W, k}> P_{out_S,k}When, energy-storage system of accumulator absorbs power；Work as P_{Out_W, k <}P_{out_S,k,}When, energy-storage system of accumulator Delivered power, value size are to stabilize the difference of output and wind power.

6. wind-powered electricity generation field control method is accessed based on the batteries to store energy system for improving exponential smoothing as claimed in claim 5, It is characterized in, the target output of power and system that energy-storage system of accumulator needs to absorb or discharge calculates, it specifically includes:

The power data of wind power plant a certain period is inputted, power swing rate is calculated and needs energy-storage system of accumulator to stabilize fluctuation Output or absorb performance number；

Output power fluctuation of wind farm after considering the output power of energy-storage system of accumulator, and whether judge stability bandwidth at this time Meet the requirement of professional standard table, if not satisfied, the charge-discharge electric power of adjustment energy-storage system of accumulator, if satisfied, calculating this When battery state-of-charge；

Judge whether storage battery charge state meets the requirements, if not satisfied, directly terminating；If satisfied, carrying out next step iteration, directly To termination condition is met, the output power and corresponding energy-storage system of accumulator that wind power plant is exported after iteration need to charge or put The capacity of electricity.