CN116796540A - Large photovoltaic power station energy storage capacity configuration method considering light rejection rate and prediction precision - Google Patents

Abstract

The present invention proposes a method for configuring energy storage capacity of large-scale photovoltaic power stations that considers the light abandonment rate and prediction accuracy. First, the basic data of the photovoltaic power station is deeply mined, and the annual light abandonment phenomenon of the photovoltaic power station and the basic situation of the photovoltaic prediction pass rate are analyzed. Then a method is proposed to calculate the monthly photovoltaic prediction qualification rate and penalty cost; then the specific algorithm process of energy storage capacity configuration is given, and the simulation results are obtained by running the original data. Through the light abandonment rate, predicted qualification rate and economic A comprehensive configuration plan for energy storage capacity can be obtained by comparing the specific trade-offs. The present invention provides a practical energy storage configuration technical solution for photovoltaic power stations, and realizes automatic analysis of photovoltaic power station configuration energy storage optimization solutions.

Description

Energy storage capacity allocation method for large-scale photovoltaic power stations considering light abandonment rate and prediction accuracy

Technical field

The invention relates to the field of energy storage capacity configuration of new energy power stations, specifically a method for configuring energy storage capacity of large-scale photovoltaic power stations that considers the light abandonment rate and prediction accuracy.

Background technique

Photovoltaic power generation has the characteristics of intermittent and fluctuation, and the power output is unstable. In addition, the grid dispatch center uses the predicted power of photovoltaic power plants as its planned output, but there is still a large deviation between the predicted value and the actual value, resulting in a mismatch between the actual generated power and the planned output. This may not only cause photovoltaic power curtailment This phenomenon may also cause insufficient system backup, leading to safety and reliability problems. In addition, additional expenditures will be increased for power stations facing assessment problems. Due to its flexible charging and discharging characteristics, energy storage plays an indispensable role in smoothing the power fluctuations of new energy sources, promoting a high proportion of green energy to the grid, and reducing power prediction deviations. It is an important means to solve the problem of photovoltaic power station grid connection. , while ensuring economic benefits while promoting energy conservation and emission reduction. However, the initial investment cost of energy storage is high, so how to conduct reasonable capacity allocation is of great research significance.

In summary, there have been many studies on optical storage capacity allocation methods in recent years, but there are very few optical storage allocation calculation models that consider the light abandonment rate and prediction deviation. From the perspective of the power station, the optimal allocation of energy storage considering the technical and economical aspects can not only improve the photovoltaic power station Its own operating characteristics can also indirectly improve the reliability and environmental protection of system operation, which can be said to provide a theoretical basis for the effective allocation of energy storage on the photovoltaic power station side.

Contents of the invention

In order to solve the above problems, the present invention provides a large-scale photovoltaic power station energy storage capacity configuration method that considers the light abandonment rate and prediction accuracy. Through the specific algorithm flow of energy storage capacity configuration given by the present invention and through running historical data simulation, the light abandonment can be calculated The comprehensive configuration scheme of energy storage capacity is obtained by weighing the efficiency, predicted qualification rate and economics, which provides a practical energy storage configuration technical scheme for photovoltaic power stations, and further verifies the feasibility of configuring energy storage on the power station side. Solve the problems raised in the above background technology.

In order to solve the above technical problems, the technical solutions adopted by the present invention are:

According to the detailed assessment rules of Jiangsu, the present invention proposes a comprehensive configuration plan that takes into account assessment indicators and energy storage economy from the perspective of a power station. It performs automatic extraction and analysis of assessment indicators based on the annual operation data of large-scale photovoltaic power stations. Combined with the 4000MW photovoltaic The power station conducted a case analysis, starting from a large-scale photovoltaic energy storage configuration calculation model that improves the light abandonment rate and prediction accuracy, and through multi-scenario simulation analysis. The results show that the comprehensive benefits are optimal when the energy storage capacity is configured to be 11% of the photovoltaic installed capacity, effectively reducing The penalty cost of photovoltaic power plants. Therefore, the present invention's optimal plan for configuring energy storage for a specific photovoltaic power station should be combined with the analysis of the photovoltaic power station's annual operation characteristics and specific assessment rules, which is conducive to improving the technical and economic efficiency of the joint operation of photovoltaic power station energy storage.

A method for configuring energy storage capacity of large-scale photovoltaic power stations that considers the light abandonment rate and prediction accuracy, including the extraction of photovoltaic power station characteristic indicators and statistical analysis, thereby measuring the light abandonment rate and prediction pass rate of the power station's historical data, and then establishing assessment indicators that take into account and economical energy storage capacity allocation methods.

A method for configuring energy storage capacity of large-scale photovoltaic power stations that considers light abandonment rate and prediction accuracy, including the following steps:

Step 1: Establish characteristic indicators of photovoltaic power station’s light abandonment rate and predicted qualification rate;

Step 2: Enter basic data and set parameters. Statistics of monthly photovoltaic forecast pass rates and calculation of penalty costs;

Step 3: Power constraints are used to filter out the abandoned optical power data set that can be absorbed by energy storage;

Step 4: Energy constraints to avoid the loss of battery life due to full discharge, and the battery charge and discharge capacity is stipulated to be 90% of the rated capacity;

Step 5: Output the simulation results, including technical indicators such as energy storage net income C, light abandonment rate and loss cost;

As a further improvement of the present invention, step 1 establishes and calculates characteristic indicators such as the light abandonment rate of the photovoltaic power station. The specific steps are as follows:

Step 1-1: Through statistics and analysis of photovoltaic power station sample data sets, the formula for measuring the light abandonment power of each sample point is:

p _s,i =max(0,p _a,i -p _c,i )#(1)

p _s,i is the abandoned light power of the i-th sampling point, p _a,i and p _c,i are the actual power and predicted power of the i-th sampling point respectively. When p _a,i <p _c,i , All photovoltaic power generation will be used to respond to grid dispatch. If p _a,i > p _c,i , there will be surplus photovoltaic power generation, which will lead to a certain amount of light abandonment.

Step 1-2: Establish a model to calculate the total monthly light abandonment power and light abandonment rate of the photovoltaic power station;

p _s,m is the monthly total light abandonment power data set, d(m) represents the total number of days in the mth month, and λ(m) represents the monthly light abandonment rate data set.

Step 1-3: Establish a prediction qualification rate model for photovoltaic power plants;

p _a,i is the actual power of the i-th sample point, p _c,i is the predicted power of the i-th sample point, C _ap is the rated capacity, combined with the "Jiangsu Province Electric Power Grid Connection Operation Management Rules" on the power prediction. The qualified points are counted and the penalty cost caused by the forecast deviation is calculated every month.

As a further improvement of the present invention, step 2 is based on the basic operation data of the photovoltaic power station, and then counts the monthly photovoltaic prediction pass rate and calculates the penalty cost. The specific steps are as follows:

Step 2-1: Input data and calculate the predicted power deviation of each point through equation (1);

Step 2-2: Calculate the total number of sampling points per month, determine the specific data range corresponding to the monthly sampling set, and calculate it through Equation (5):

S(m) represents the specific data range corresponding to the data of month m in the annual data set, d(m) represents the number of days in month m, if and only if m=1,

Step 2-3: Count the total number of unqualified points every month, and find the proportion of unqualified points every month. Calculate it through equations (6) and (7):

N(m)＝{α[S(m)]<90%}#(6)

N(m) is the total number of unqualified points in month m, and α[S(m)] represents the pass rate of each point in month m.

It is the proportion of monthly unqualified points.

Step 2-4: Determine whether the proportion of monthly unqualified points is greater than 2%. If it is greater than 2%, calculate the penalty cost caused by the forecast deviation in the current month through Equation (8). If it is less than or equal to 2%, then the penalty cost in the current month is is 0.

Let: |p _a,i -p _c,i |=p _p,i , then:

p _p,i represents the absolute value of the power prediction deviation at the i-th sample point, and C _d (m) represents the penalty cost caused by the scheduling deviation in month m.

Step 2-5: Determine whether the monthly statistics have been completed, otherwise return to step 1-2;

Step 2-6: Output the calculation results such as the total number of monthly failure points, failure rate, and penalty costs for each month.

As a further improvement of the present invention, step 3 performs power constraints during the data running process, thereby screening out the abandoned optical power data set that can be absorbed by energy storage. The specific steps are as follows:

Step 3-1: Calculate the light abandonment power of each sample point through equation (1), and form a light abandonment data set corresponding to one time point;

Step 3-2: Filter out the abandoned optical power data set that can be absorbed by energy storage through equation (9). If the abandoned optical power at a certain moment is greater than the rated power of energy storage, it cannot be absorbed. The default is 0, and the update can be Power data set of energy storage and consumption;

It is a data set of abandoned light power that can be absorbed by energy storage after screening.

As a further improvement of the present invention, in step 4, in order to avoid the loss of battery life due to full discharge, the energy of the energy storage during the charge and discharge process is constrained. The specific steps are as follows:

Step 4-1: Calculate the total amount of energy storage operation in one day through equation (10):

E _T,j is the total amount of light waste that the energy storage battery has absorbed up to the j sampling point on the T day.

Step 4-2: In order to avoid the loss of battery life due to full discharge, the battery charge and discharge capacity is stipulated to be 90% of the rated capacity;

E _T,j =0.9*E _bat #(11)

Step 4-3: After the energy storage battery is fully charged during the day, it has reached a saturated state and can no longer absorb the photovoltaic waste power. Therefore, the real-time data set of the battery's waste power consumption is updated again:

As a further improvement of the present invention, step 5 outputs the simulation results, and the output includes technical indicators such as energy storage net income C, light abandonment rate, and loss cost.

E _bat is the energy storage capacity, p _cs is the energy storage rated power, C _bat is the unit capacity cost, C _PCS is the unit power cost, and C _pv,g is the photovoltaic on-grid electricity price.

As a further improvement of the present invention, in step 5, the parameter settings are as shown in Table 1 and Table 2.

Table 1 Capacity ratio settings

Table 2 Main parameter settings

Beneficial effects of the present invention:

Compared with the existing technology, the beneficial effects of the present invention: This patent proposes a large-scale photovoltaic power station energy storage capacity configuration method that considers the light rejection rate and prediction accuracy. This method has the following advantages:

1. According to the Jiangsu assessment rules, the present invention proposes a comprehensive configuration plan that takes into account assessment indicators and energy storage economy from the perspective of a power station;

2. The present invention obtains the optimal energy storage configuration plan of the photovoltaic power station by exploring the impact of the energy storage configuration plan on the light abandonment rate, photovoltaic power abandonment and energy storage economy;

3. The present invention is suitable for analyzing the operating characteristics of large-scale photovoltaic power stations within a year-round time scale, and realizes automatic analysis of photovoltaic power station configuration energy storage optimization plans;

4. Through simulation analysis, the present invention shows that when the energy storage capacity is configured as 11% of the photovoltaic installed capacity, the comprehensive benefits are optimal, the highest annual net income from energy storage can be achieved, and the penalty cost of the photovoltaic power station can be effectively reduced.

Description of the drawings

Figure 1 is a flow chart of the method proposed by the present invention.

Figure 2 is a trend chart of energy storage net income.

Figure 3 is a trend chart of the light abandonment rate and energy storage net income.

Figure 4 is a trend chart of the annual predicted failure points and penalty costs of photovoltaic power plants.

Detailed ways

A method for configuring energy storage capacity of a large-scale photovoltaic power station taking into account the light rejection rate and prediction accuracy of the present invention will be further described in detail below in conjunction with the accompanying drawings and specific implementation methods.

As shown in Figure 1, the specific implementation process of the present invention is given, a large-scale photovoltaic power station energy storage capacity configuration method that considers the light abandonment rate and prediction accuracy, which specifically includes the following steps:

Step 1: Establish characteristic indicators of photovoltaic power station’s light abandonment rate and predicted qualification rate;

Step 1-1: Through statistics and analysis of photovoltaic power station sample data sets, the formula for measuring the light abandonment power of each sample point is:

p _s,i =max(0,p _a,i -p _c,i )#(1)

p _s,i is the abandoned light power of the i-th sampling point, p _a,i and p _c,i are the actual power and predicted power of the i-th sampling point respectively. When p _a,i <p _c,i , All photovoltaic power generation will be used to respond to grid dispatch. If p _a,i > p _c,i , there will be surplus photovoltaic power generation, which will lead to a certain amount of light abandonment.

Step 1-2: Establish a model to calculate the total monthly light abandonment power and light abandonment rate of the photovoltaic power station;

p _s,m is the monthly total light abandonment power data set, d(m) represents the total number of days in the mth month, and λ(m) represents the monthly light abandonment rate data set.

Step 1-3: Establish a prediction qualification rate model for photovoltaic power plants;

p _a,i is the actual power of the i-th sample point, p _c,i is the predicted power of the i-th sample point, C _ap is the rated capacity, combined with the "Jiangsu Province Electric Power Grid Connection Operation Management Rules" on the power prediction. The qualified points are counted and the penalty cost caused by the forecast deviation is calculated every month.

Step 2: Enter basic data and set parameters. Statistics of monthly photovoltaic forecast pass rates and calculation of penalty costs;

Step 2-1: Input data and calculate the predicted power deviation of each point through equation (1);

Step 2-2: Calculate the total number of sampling points per month, determine the specific data range corresponding to the monthly sampling set, and calculate it through Equation (5):

S(m) represents the specific data range corresponding to the data of month m in the annual data set, d(m) represents the number of days in month m, if and only if m=1,

Step 2-3: Count the total number of unqualified points every month, and find the proportion of unqualified points every month. Calculate it through equations (6) and (7):

N(m)＝{α[S(m)]<90%}#(6)

N(m) is the total number of unqualified points in month m, and α[S(m)] represents the pass rate of each point in month m.

It is the proportion of monthly unqualified points.

Step 2-4: Determine whether the proportion of monthly unqualified points is greater than 2%. If it is greater than 2%, calculate the penalty cost caused by the forecast deviation in the current month through Equation (8). If it is less than or equal to 2%, then the penalty cost in the current month is is 0.

Let: |p _a,i -p _c,i |=p _p,i , then:

p _p,i represents the absolute value of the power prediction deviation at the i-th sample point, and C _d (m) represents the penalty cost caused by the scheduling deviation in month m.

Step 2-5: Determine whether the monthly statistics have been completed, otherwise return to step 1-2;

Step 2-6: Output the calculation results such as the total number of monthly failure points, failure rate, and penalty costs for each month.

Step 3: Power constraints are used to filter out the abandoned optical power data set that can be absorbed by energy storage;

Step 3-1: Calculate the light abandonment power of each sample point through equation (1), and form a light abandonment data set corresponding to one time point;

Step 3-2: Filter out the abandoned optical power data set that can be absorbed by energy storage through equation (9). If the abandoned optical power at a certain moment is greater than the rated power of energy storage, it cannot be absorbed. The default is 0, and the update can be Power data set of energy storage and consumption;

It is a data set of abandoned light power that can be absorbed by energy storage after screening.

Step 4: Energy constraints to avoid the loss of battery life due to full discharge, and the battery charge and discharge capacity is stipulated to be 90% of the rated capacity;

Step 4-1: Calculate the total amount of energy storage operation in one day through equation (10):

E _T,j is the total amount of light waste that the energy storage battery has absorbed up to the j sampling point on the T day.

Step 4-2: In order to avoid the loss of battery life due to full discharge, the battery charge and discharge capacity is stipulated to be 90% of the rated capacity;

E _T,j =0.9*E _bat #(11)

Step 4-3: After the energy storage battery is fully charged during the day, it has reached a saturated state and can no longer absorb the photovoltaic waste power. Therefore, the real-time data set of the battery's waste power consumption is updated again:

Step 5: Output the simulation results, including technical indicators such as energy storage net income C, light abandonment rate and loss cost;

E _bat is the energy storage capacity, p _cs is the energy storage rated power, C _bat is the unit capacity cost, C _PCS is the unit power cost, and C _pv,g is the photovoltaic on-grid electricity price.

In step 5, the parameter settings are as shown in Table 1 and Table 2.

Table 1 Capacity ratio settings

Table 2 Main parameter settings

In order to evaluate the feasibility and effectiveness of the present invention, the present invention takes the basic data of 35,040 sample points with a sampling interval of 15 minutes from the Belgian photovoltaic power station as an example to study the configuration of energy storage capacity. Among them, the capacity of the photovoltaic power station is 4000MW. According to the regulations on energy storage capacity configuration in the document No. 949 of Su Fagai Energy Development {2021}: areas south of the Yangtze River are equipped with a power ratio of 8% and above (duration is two hours, the same below) ); In principle, the area north of the Yangtze River is equipped with a power ratio of 10% and above. Based on the above regulations and combined with basic data, the energy storage capacity ratio is set as shown in Table 1 above.

According to the parameter settings in Table 1, the calculation example of the present invention mainly consists of 6 scenarios, whose energy storage capacity ratios are 8%, 10%, 11%, 12%, 13% and 15% respectively. By calculating the different ratios Rated power size, and then determine the capacity under each ratio. Among them, the energy storage system is considered based on a daily cycle, that is, when the photovoltaic power station generates more power than the dispatch plan, the remaining power is used to charge the energy storage, and the energy storage system starts to discharge at night. Through joint optimized operation, the photovoltaic storage system can not only reduce the light abandonment rate, but also improve the absorption capacity and prediction accuracy of the photovoltaic power station. In the above energy storage capacity allocation method based on assessment indicators and economics, the basic parameter settings in step 5 are as shown in Table 2 above.

In order to count the total number of unqualified points every month from a huge data set and calculate the penalty cost, a method is proposed above to count the monthly photovoltaic prediction pass rate and calculate the penalty cost. Step 1- is completed through calculation. The output results after 2 are shown in Table 3:

Table 3 calculation results

From the calculation results in Table 3, it can be concluded that for 8 months in 2021, if the predicted power assessment fails, the penalty fee to be paid totals about 1.0062 million yuan, and if the assessment is passed in 4 months, no penalty fee is required. This seriously affects Normal dispatch of the power system and causing additional expenditures for photovoltaic power plants. In summary, the above problems can be solved by configuring energy storage in photovoltaic power stations. Based on the above optical storage capacity configuration calculation method, simulate and calculate the energy storage configuration of large photovoltaic power stations, complete steps 1 to 5, and the calculation results are shown in Table 4:

Table 4 Simulation results

In order to more clearly compare the influence of energy storage configuration on various indicators, the following is a detailed analysis of the simulation results based on the established evaluation indicators, mainly including the impact of energy storage capacity on the net cost of the power station, the light abandonment rate and the prediction qualification rate.

(1) From the perspective of energy storage net income, as the energy storage configuration capacity increases, the net energy storage income shows a trend of first increasing and then decreasing, as shown in Figure 2, because energy storage is charged when photovoltaic output is excess. , discharging at night, a certain amount of income was obtained, with the annual net income reaching the maximum of 3.17 million. However, when the optical storage capacity ratio exceeds 11%, the income begins to decline, and when the ratio exceeds 12%, the income is even negative.

(2) When the optical storage capacity is configured at the minimum ratio of 8%, the annual light abandonment rate is already less than 5%, which is in line with the standards of the "Clean Energy Consumption Action Plan" formulated by the country. With the increase in energy storage configuration capacity , the light abandonment rate continues to decrease, from the highest 4.32% to 1.5%, and the photovoltaic utilization rate is greatly improved, as shown in Figure 3. However, with the large-scale deployment of energy storage, although photovoltaic consumption has been improved to a certain extent However, when the ratio exceeds 12%, the energy storage income is negative. Therefore, reducing the light abandonment rate requires comprehensive consideration of the economics of energy storage.

(3) As the proportion of optical storage capacity increases, the annual predicted failure points and penalty costs change in the same trend. The larger the energy storage configuration capacity, the penalty cost and the annual predicted failure points continue to decrease, as shown in Figure 4. When the energy storage capacity ratio is 8% and 10%, the penalty cost remains unchanged and the annual predicted failure points do not change, proving that the allocated capacity is too small.

Based on the analysis results of the above indicators, when the energy storage capacity ratio is 11%, that is, 440MW/880MWh, 2h, the economy is the best, with the maximum annual net income reaching 3.17 million, and the predicted power deviation is significantly improved, and the light abandonment rate Only 2.89%, in line with relevant regulations and standards. While improving the economy and photovoltaic consumption rate, the prediction deviation is improved, the prediction qualification rate is improved, and the penalty cost is reduced. Therefore, when the optical-to-storage capacity ratio is 11%, the overall benefit is the best.

According to the Jiangsu assessment rules, the present invention proposes a comprehensive configuration plan that takes into account assessment indicators and energy storage economy from the perspective of a power station. It performs automatic extraction and analysis of assessment indicators based on the annual operation data of large-scale photovoltaic power stations, and combines it with a 4000MW photovoltaic power station. Case analysis, from the large-scale photovoltaic energy storage configuration calculation model to improve the light abandonment rate and prediction accuracy, through multi-scenario simulation analysis, the results show that the comprehensive benefits are optimal when the energy storage capacity is configured to be 11% of the photovoltaic installed capacity, and energy storage can be achieved The annual net income is the highest. At this time, the light abandonment rate drops to 2.89%. At the same time, the number of unqualified photovoltaic forecast points is reduced by 203, which effectively reduces the penalty cost of the photovoltaic power station. Therefore, the present invention's optimal plan for configuring energy storage for a specific photovoltaic power station should be combined with the analysis of the photovoltaic power station's annual operation characteristics and specific assessment rules, which is conducive to improving the technical and economic efficiency of the joint operation of photovoltaic power station energy storage.

The above are only the preferred embodiments of the present invention. It should be pointed out that those of ordinary skill in the art can make several improvements and modifications without departing from the principles of the present invention. These improvements and modifications can also be made. should be regarded as the protection scope of the present invention.

Claims (7)

1. The large-scale photovoltaic power station energy storage capacity configuration method is characterized in that characteristic indexes of the photovoltaic power station are extracted and statistically analyzed, so that the light rejection rate and the prediction qualification rate of historical data of the power station are measured and calculated, and the energy storage capacity configuration method considering assessment indexes and economy is further established.

2. The method for configuring the energy storage capacity of the large photovoltaic power station taking the light rejection rate and the prediction accuracy into consideration as claimed in claim 1, comprising the following steps:

step 1: establishing the characteristics indexes of the light rejection rate and the predicted qualification rate of the photovoltaic power station;

step 2: inputting basic data and setting parameters; counting photovoltaic prediction qualification rate per month and calculating punishment cost;

step 3: power constraint, so as to screen out a waste light power data set which can be consumed by energy storage;

step 4: energy constraint, avoiding the loss of full charge and discharge to the service life of the battery, and prescribing the charge and discharge capacity of the battery to be 90% of rated capacity;

step 5: and outputting a simulation result, and outputting technical indexes including energy storage net benefit C, light rejection rate and loss cost.

3. The method for configuring the energy storage capacity of the large-scale photovoltaic power station taking the light rejection rate and the prediction accuracy into consideration as claimed in claim 2, wherein the step 1 is to build and calculate the characteristic indexes of the light rejection rate and the prediction qualification rate of the photovoltaic power station, and comprises the following specific steps:

step 1-1: through statistics and analysis of a photovoltaic power station sample data set, a formula for measuring and calculating the light rejection power of each sample point is as follows:

p _s,i ＝max(0,p _a,i -p _c,i )#(1)

p _s,i optical power p is discarded for the ith sampling point _a,i And p _c,i The actual power and the predicted power of the ith sampling point are respectively, when p _a,i <p _c,i At this time, the photovoltaic power generation power will be fully used for responding to the power grid dispatching, if p _a,i >p _c,i When the photovoltaic power generation power remains, a certain light rejection phenomenon can be caused;

step 1-2: establishing and calculating the total monthly light rejection power and the light rejection rate model of the photovoltaic power station;

p _s,m for a total monthly light rejection power dataset, d (m) represents the total number of days of month m, λ (m) represents the light rejection rate dataset for each month;

step 1-3: establishing a prediction qualification rate model of the photovoltaic power station;

p _a,i for the actual power of the ith sample point, p _c,i Predicted power for the ith sample point, C _ap And counting unqualified points of power prediction for rated capacity, and calculating punishment cost of each month caused by prediction deviation.

4. The method for configuring the energy storage capacity of the large-scale photovoltaic power station taking the light rejection rate and the prediction accuracy into consideration as claimed in claim 3, wherein the step 2 is based on the basic operation data of the photovoltaic power station, and further comprises the following specific steps of:

step 2-1: inputting data, and calculating the predicted power deviation of each point through the formula (1);

step 2-2: calculating the total sampling point number of each month, determining the specific data range corresponding to the sampling set of each month, and calculating by the formula (5):

s (m) represents the specific data range to which the data of month m corresponds in the annual data set, d (m) represents how many days the month m shares, if and only if m=1,

step 2-3: counting the total number of disqualified points in each month, and obtaining the proportion of disqualified points in each month, wherein the proportion is calculated by the formulas (6) and (7):

N(m)＝{α[S(m)]<90％}#(6)

n (m) is the total number of unqualified points in the month of m, and alpha [ S (m) ] represents the qualification rate of each point in the month of m;

the proportion of the number of unqualified points in the month;

step 2-4: judging whether the proportion of the disqualified points in each month is more than 2%, if so, calculating punishment cost caused by prediction deviation in the current month through a formula (8), and if so, setting punishment cost in the current month to be 0;

and (3) making: p _a,i -p _c,i |＝p _p,i Then:

p _p,i representing the absolute value of the i-th sample point power prediction bias, C _d (m) represents penalty cost due to scheduling bias for m months;

step 2-5, judging whether the monthly conditions are counted completely, otherwise, returning to the step 1-2;

step 2-6: and outputting the calculation results of the total number of the unqualified points of the month, the unqualified rate, the punishment cost of each month and the like.

5. The method for configuring the energy storage capacity of the large photovoltaic power station taking the light rejection rate and the prediction accuracy into consideration as set forth in claim 4, wherein the step 3 is to perform power constraint in the data operation process, thereby screening out the light rejection power data set which can be consumed by the stored energy, and specifically comprises the following steps:

step 3-1: calculating the light rejection power of each sample point by the formula (1), and forming a light rejection data set corresponding to the time points one by one;

step 3-2: screening out a waste light power data set which can be consumed by energy storage through a formula (9), if the waste light power at a certain moment is larger than the rated power of the energy storage and cannot be absorbed, defaulting to 0, and updating the power data set which can be consumed by the energy storage;

and the data set is a discarded light power data set which can be consumed by stored energy after screening.

6. The method for configuring the energy storage capacity of the large-scale photovoltaic power station taking the light rejection rate and the prediction accuracy into consideration as claimed in claim 5, wherein in step 4, in order to avoid the loss of the battery life caused by full charge and discharge, the energy of the stored energy in the charge and discharge process is constrained, and the specific steps are as follows:

step 4-1: calculating the total electric quantity of the energy storage operation in one day by the formula (10):

E _T,j the total amount of waste light consumed by the energy storage battery from the sampling point of j to the sampling point of j on the T th day;

step 4-2: in order to avoid the loss of the full charge and discharge to the service life of the battery, the charge and discharge capacity of the battery is specified to be 90% of the rated capacity;

E _T,j ＝0.9*E _bat #(11)

step 4-3: after the energy storage battery is fully charged in the course of a day, the energy storage battery reaches a saturated state, and the photovoltaic waste light electric quantity is not consumed any more, so that the real-time data set of the consumed waste light power of the battery is updated again:

7. the method for configuring the energy storage capacity of the large photovoltaic power station with consideration of the light rejection rate and the prediction accuracy according to claim 6, wherein after all data are circulated in 365 days in one year, step 5 outputs the net energy storage benefit C:

E _bat is of energy storage capacity, p _cs Rated power for energy storage, C _bat Cost per unit volume, C _PCS Is the unit power cost, C _pv,g Photovoltaic internet electricity price.