A method and device for configuring reserve capacity for a wind-solar-storage combined system

Abstract

The application faces a reserve capacity configuration method of a wind-light-storage combined system, comprising: generating typical scenarios of new energy predicted output clustering based on an improved K-means clustering algorithm according to new energy predicted output data of a wind-light-storage combined system in a certain region; determining the new energy predicted output reserve safety inclusion coefficient of the typical scenarios through a calculation method of the new energy predicted output reserve safety inclusion coefficient and a power supply loss risk rate constraint; calculating the new energy reserve safety inclusion coefficient of the wind-light-storage combined system based on the pumped storage coordination according to the new energy predicted output data and the pumped storage reservoir capacity of the next period, and solving the reserve configuration capacity of the new energy output prediction clustering. The application solves the problem of over-conservative new energy reserve inclusion caused by the volatility and uncertainty of new energy output, further reduces the starting mode of conventional thermal power units, and improves the new energy consumption space.

Description

Technical Field

[0001] This invention belongs to the field of power system operation and control technology, specifically, it relates to a method and equipment for configuring standby capacity for wind-solar-storage integrated systems. Background Technology

[0002] New energy sources, represented by wind and solar power, exhibit strong randomness and anti-peak-shaving characteristics, leading to increasingly prominent issues such as insufficient systemic peak-shaving and reserve requirements, posing a series of challenges to the safe and stable operation of the power grid. While large-scale grid connection of new energy sources can reduce the generation costs of thermal power, ensuring the safe and stable operation of the power system often requires reserving sufficient reserve capacity to cope with the volatility and uncertainty of new energy sources, which in turn increases the system's operating costs. As the penetration rate of new energy sources continues to increase, when the peak-shaving capacity of thermal power is limited and unable to smooth out the system peak-valley difference caused by fluctuations in new energy output and its anti-peak-shaving characteristics, severe curtailment of new energy sources may occur.

[0003] Currently, the spinning reserve capacity in the power system is provided by conventional units. In fact, renewable energy generation can also participate in operation as a backup power source. Furthermore, the combustion of fossil fuels in thermal power plants is a significant contributor to environmental pollution. In the current context of energy conservation and emission reduction, simply optimizing system operation economics is no longer sufficient to meet societal demands and power system operating standards. Therefore, for modern power systems containing a high proportion of renewable energy, considering the uncertainty of renewable energy output and optimizing the output of various power sources within the system is of profound significance for reducing system operating costs, increasing the absorption of renewable energy, and mitigating environmental problems. Summary of the Invention

[0004] In view of this, the present invention provides a method and equipment for configuring reserve capacity for wind-solar-storage combined systems, in order to solve the problem that the inclusion of new energy reserves is too conservative due to the volatility and uncertainty of new energy output, further reduce the start-up mode of conventional thermal power units, and increase the space for new energy consumption.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: a method for configuring standby capacity for wind-solar-storage integrated systems, specifically including the following steps:

[0006] Step 1: Based on the predicted output data of the wind-solar-storage combined system in a certain region, determine the optimal number of typical scenarios based on the improved K-means clustering algorithm, and generate typical scenarios for the predicted output clustering of new energy.

[0007] Step 2: Using the calculation method of the new energy predicted output reserve safety inclusion coefficient and the power supply loss risk rate constraint, determine the new energy predicted output reserve safety inclusion coefficient for typical scenarios.

[0008] Step 3: Based on the predicted output data of new energy sources and the capacity of pumped storage reservoirs for the next time period, calculate the safety inclusion coefficient of new energy reserve for the wind-solar-storage combined system based on pumped storage coordination, and solve for the reserve configuration capacity of the predicted cluster of new energy output.

[0009] Furthermore, step 1 includes the following sub-steps:

[0010] Step 101: Collect actual power output data P of new energy sources in a certain region at different times. Xn ={p Xn,h} and the corresponding new energy power output forecast data P X * n ={p X * n,h}, where P Xn Let p be a sample set of actual power output data from new energy sources for a certain period on day n. Xn,h For the sample of actual power output data of new energy at the h-th sampling point during a certain time period on day n, P X * n p is the sample set of predicted power output data for new energy sources corresponding to the nth day. X * n,h This is a sample of the predicted power output data of new energy sources at the h-th sampling point corresponding to the nth day;

[0011] Step 102: Set the initial number of typical clustering scenarios and specify the cluster center for each typical clustering scenario;

[0012] Step 103: Calculate the predicted power output data P of new energy sources using the K-means clustering algorithm. X * n The similarity coefficients to each cluster center are used to determine the predicted power output data P of the new energy sources with the lowest similarity coefficients. X * n They are categorized into the corresponding typical clustering scenarios;

[0013] The predicted power output data P of the new energy source X * n The similarity coefficient d(P) to each cluster center X * n O j The calculation process for ) is as follows:

[0014]

[0015] Among them, O jLet j be the cluster center for the j-th typical clustering scenario;

[0016] Step 104: Calculate the mean of all new energy predicted output data samples in the same cluster typical scenario, and take the new energy predicted output data sample that is closest to the mean as the new cluster center of the cluster typical scenario;

[0017] Step 105: If the new cluster centers of the typical clustering scenarios change, repeat the clustering process of steps 103-104 according to the new cluster centers; otherwise, determine whether the number of typical clustering scenarios is greater than 2. If so, decrement the number of clustering scenarios by 1, specify the cluster centers of each typical clustering scenario, and execute step 103. Otherwise, calculate the average profile coefficient Z of all typical clustering scenarios based on the median, select the number of typical clustering scenarios with the largest Z as the optimal value of the number of typical clustering scenarios, and generate typical scenarios for new energy power output prediction clustering.

[0018] Furthermore, the calculation process for the median-based average silhouette coefficient Z is as follows:

[0019]

[0020] Among them, u i The sample P represents the predicted output data of new energy sources. X * i And all other new energy predicted power output data samples p belonging to the typical clustering scenario D x * j The median distance between them, u i =median[d(p X * i ,p X * j )];v i The sample P represents the predicted output data of new energy sources. X * i And all new energy predicted output data samples P in the nearest neighbor clustering typical scenario E that do not belong to clustering typical scenario D X * k The median of the distances between them. c represents the number of new energy power output prediction data samples in cluster scenario D.

[0021] Furthermore, step 2 includes the following sub-steps:

[0022] Step 201: Cluster typical scenarios D of new energy predicted output. k (k = 1, 2, ..., K) UThe predicted output data sample set of new energy sources in the data is combined with the corresponding actual output data sample set of new energy sources and standardized to obtain the prediction compliance rate of the predicted output data sample set of new energy sources:

[0023]

[0024] Among them, X m L Typical scenario D for clustering to contribute to new energy prediction k The m-th new energy source predicted power output data sample set X m The predicted compliance rate of new energy output, where H is the total amount of samples taken over the time interval, h is the index of H, and x m,h For X m The new energy forecast compliance rate of the h-th new energy power output data sample. p X * n,h For the new energy power output data sample of the h-th sampling point during a certain period on day n, p Xn,h This is a sample of actual power output data of new energy sources at the h-th sampling point during a certain time period on day n.

[0025] Step 202: Set B% as the maximum allowable power outage risk rate for the wind-solar-storage combined system, and calculate the maximum power outage risk rate B% ​​in typical scenario D. k The quantile F of the predicted achievement rate in the ascending order of the new energy power output forecast data sample set. B％ =max[floor(M×B%),1], according to quantile F B％ Index the data to find the Xm corresponding to the power outage risk rate B% ​​for each time period. L The backup safety inclusion coefficient P, which serves as a typical scenario for the clustering, is the predicted output of new energy sources. X *L (B%), where M is the m-th new energy predicted output data sample set X. m The total number of samples in the data, where floor() is the floor function.

[0026] Furthermore, step 3 includes the following sub-steps:

[0027] Step 301: Obtain the predicted output data of the wind-solar-storage combined system and the pumped storage reservoir capacity for the next time period;

[0028] Step 302: Calculate the similarity coefficient between the predicted renewable energy output data for the next time period and the cluster centers of typical scenarios in the predicted renewable energy output clustering. Select the typical scenario with the smallest similarity coefficient as the predicted renewable energy output scenario for the time period, and obtain the backup safety inclusion coefficient P of the predicted renewable energy output under the power supply failure risk rate B%.X *L (B%)

[0029] Step 303: Obtain the pumped storage status based on the pumped storage reservoir capacity for the next time period, and determine the pumped storage coordination coefficient P by combining it with the predicted maximum output of new energy sources for the next time period. C L :

[0030]

[0031] Among them, P C It is in pumped storage mode. q is the pumped storage power output coefficient. Q C Q represents the reservoir capacity of the pumped storage upper reservoir of the wind-solar-storage combined system. Cmin Q represents the minimum reservoir capacity of a pumped-storage hydroelectric system for a combined wind-solar-storage system. C ^ To provide the reservoir capacity required for the pumped storage of the wind-solar-storage combined system to maintain full power generation in the next period, P Cmax P represents the maximum pumped storage output of the wind-solar-storage combined system. X * max This represents the maximum predicted output of new energy sources for the next time period.

[0032] Step 304: Using the pumped storage coordination coefficient P C L And the backup safety inclusion factor P of the predicted output of new energy sources under the power supply failure risk rate B%. X *L (B%), calculate the reserve safety inclusion factor P for the predicted output of new energy sources after coordination of pumped storage. X+C *L (B%):

[0033] P X+C *L (B%)=P X *L (B%)+P C L

[0034] Step 305: Based on the reserve safety inclusion factor P of the predicted output of new energy sources after pumped storage coordination. X+C *L (B%) and the minimum predicted output of new energy sources in the next period are used to obtain the reserve capacity S of the new energy output prediction cluster:

[0035] S = P X+C *L (B%)·P X* min .

[0036] Furthermore, the present invention also provides a computer-readable storage medium storing a computer program that causes a computer to execute the described method for configuring standby capacity for wind-solar-storage combined systems.

[0037] Furthermore, the present invention also provides an electronic device comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, it implements the aforementioned method for configuring standby capacity for wind-solar-storage combined systems.

[0038] Compared with the prior art, the beneficial effects of the present invention are:

[0039] 1. The backup capacity configuration method for wind-solar-storage combined systems of the present invention uses an improved K-means clustering algorithm to determine the number of typical clustering scenarios based on the median average profile coefficient and generate typical scenarios for new energy predicted output clustering. Since the evaluation profile coefficient based on the median is not easily affected by individual extreme values ​​within the scenario, the generated typical scenarios for new energy predicted output clustering are more representative.

[0040] 2. The backup capacity configuration method for wind-solar-storage combined systems of the present invention determines the backup safety inclusion coefficient of new energy predicted output in typical scenarios by calculating the backup safety inclusion coefficient of new energy predicted output and constraining the power supply loss risk rate. It plans the proportion of new energy included in the backup based on historical output, laying the foundation for the rapid calculation of the proportion coefficient of new energy included in the backup in future periods.

[0041] 3. The backup capacity configuration method for wind-solar-storage integrated systems of this invention calculates the backup safety inclusion coefficient of the wind-solar-storage integrated system based on the predicted output data of new energy sources in the next period and the capacity of pumped storage reservoirs, and solves for the backup configuration capacity of new energy output prediction clusters. When the pumped storage units are added to make them able to quickly act as generator units to supplement the load power supply when there is a certain load power supply gap due to the excessive proportion of new energy included in the backup, the coordination capability is improved, thereby increasing the proportion of new energy included in the backup of the system. This solves the problem of overly conservative inclusion of new energy in the backup due to the volatility and uncertainty of new energy output, further reduces the start-up mode of conventional thermal power units, and increases the space for new energy consumption. Attached Figure Description

[0042] Figure 1 This is the overall flowchart of the backup capacity configuration method for wind-solar-storage integrated systems of the present invention;

[0043] Figure 2 This is a schematic diagram illustrating the impact of incorporating new energy sources into the reserve on the absorption capacity of the wind-solar-storage combined system;

[0044] Figure 3 This is a flowchart of the clustering method for typical scenarios of new energy power output prediction in this invention;

[0045] Figure 4 This is a flowchart of the backup configuration capacity for solving the clustering of new energy output prediction in this invention. Detailed Implementation

[0046] The technical solution of the present invention will be further explained and described below with reference to the accompanying drawings.

[0047] like Figure 1 The following is a general flowchart of the backup capacity configuration method for wind-solar-storage integrated systems according to the present invention, which specifically includes the following steps:

[0048] Step 1: Based on the predicted power output data of a wind-solar-storage combined system in a certain region, and using an improved K-means clustering algorithm, determine the optimal number of typical scenarios and generate typical scenarios for the predicted power output cluster. In the improved K-means clustering algorithm of this invention, the median-based average silhouette coefficient is less susceptible to the influence of individual extreme values ​​within a scenario, making the generated typical scenarios for the predicted power output cluster more representative; specifically, as... Figure 3 Step 1 includes the following sub-steps:

[0049] Step 101: Based on the load time-series characteristics, divide the day into different time periods and collect the actual output data P of new energy sources in a certain region for each time period. Xn ={p Xn,h} and the corresponding new energy power output forecast data P X * n ={p X * n,h}, where P Xn Let p be a sample set of actual power output data from new energy sources for a certain period on day n. Xn,h For the sample of actual power output data of new energy at the h-th sampling point during a certain time period on day n, P X * n p is the sample set of predicted power output data for new energy sources corresponding to the nth day. X * n,h This is a sample of the predicted power output data of new energy sources at the h-th sampling point corresponding to the nth day;

[0050] Step 102: Set the initial number of typical clustering scenarios and specify the cluster center for each typical clustering scenario;

[0051] Step 103: Calculate the predicted power output data P of new energy sources using the K-means clustering algorithm. X * n The similarity coefficients to each cluster center are used to determine the predicted power output data P of the new energy sources with the lowest similarity coefficients. X * n They are categorized into the corresponding typical clustering scenarios;

[0052] In this invention, the predicted power output data P of new energy sources X * n The similarity coefficient d(P) to each cluster center X * n O j The calculation process for ) is as follows:

[0053]

[0054] Among them, O j Let j be the cluster center for the j-th typical clustering scenario;

[0055] Step 104: Calculate the mean of all new energy predicted output data samples in the same cluster typical scenario, and take the new energy predicted output data sample that is closest to the mean as the new cluster center of the cluster typical scenario;

[0056] Step 105: If the new cluster centers of the typical clustering scenarios change, repeat the clustering process of steps 103-104 according to the new cluster centers; otherwise, determine whether the number of typical clustering scenarios is greater than 2. If so, decrement the number of clustering scenarios by 1, specify the cluster centers of each typical clustering scenario, and execute step 103. Otherwise, calculate the average profile coefficient Z of all typical clustering scenarios based on the median, select the number of typical clustering scenarios with the largest Z as the optimal value of the number of typical clustering scenarios, and generate typical scenarios for new energy power output prediction clustering.

[0057] Traditional K-means clustering algorithms randomly select initial cluster centers and determine the number of clusters K. Once the initial cluster centers and K are determined, the algorithm calculates the entire iteration process automatically. Different initial cluster centers and K are selected each time, resulting in different clustering results, which significantly affects the stability of the clustering. Furthermore, clustering effectiveness analysis based on the average distance silhouette coefficient not only clearly shows high sample similarity within typical scenarios but also effectively shows high dissimilarity between sample points in typical clustering scenarios. However, when the sample distance difference is large, the average distance cannot accurately represent the characteristics of the sample points. The new energy output prediction data in this invention, due to its high randomness, may exhibit extreme values. The calculation of the average distance is related to each data point and is easily affected by extreme values. Any change in any data point can cause a change in the average distance, leading to unstable results. This invention uses an improved K-means clustering algorithm to determine the optimal number of clustering scenarios based on the silhouette coefficient of the median. By introducing the concept of the median, the characteristics of the sample points are better represented. The median is related to the arrangement position of the data. It is a representative value at the middle position of a set of data and is not affected by extreme values ​​of the data. This ensures that the typical scenarios of new energy power generation prediction for each time period generated by clustering are relatively stable and more representative.

[0058] The calculation process of the average profile coefficient Z based on the median in this invention is as follows:

[0059]

[0060] Among them, u i The sample P represents the predicted output data of new energy sources. X * i And all other new energy predicted power output data samples p belonging to the typical clustering scenario D x * j The median distance between them, u i =median[d(p X * i ,p X * j )];v i The sample P represents the predicted output data of new energy sources. X * i And all new energy predicted output data samples P in the nearest neighbor clustering typical scenario E that do not belong to clustering typical scenario D X * k The median of the distances between them. c represents the number of new energy power output prediction data samples in cluster scenario D.

[0061] Step 2: Using the calculation method of the renewable energy forecast output reserve safety inclusion coefficient and the power supply loss risk rate constraint, determine the renewable energy forecast output reserve safety inclusion coefficient for typical scenarios based on the renewable energy forecast output reserve safety inclusion coefficient. Under the constraint of power supply loss risk of the wind-solar-storage combined system, calculate the renewable energy safety inclusion coefficient according to different clustered typical scenarios, further reducing the minimum output value of thermal power units in the system at each time period, and improving the system's renewable energy absorption capacity, such as... Figure 2 The diagram illustrates the impact of incorporating renewable energy into the system's reserve capacity on the renewable energy absorption capacity of the wind-solar-storage combined system. The renewable energy absorption capacity of the system is primarily determined by the system's real-time load and the minimum output of thermal power units, i.e., their operating capacity. The operating capacity of thermal power units depends on the system's reserve requirements. When a portion of the predicted renewable energy output is incorporated into the system's reserve, the operating capacity of thermal power units can be reduced while maintaining the total system reserve requirements, thereby increasing the system's renewable energy absorption capacity. Specifically, this includes the following sub-steps:

[0062] Step 201: Cluster typical scenarios D of new energy predicted output. k (k = 1, 2, ..., K) U The predicted output data sample set of new energy sources in the data is combined with the corresponding actual output data sample set of new energy sources and standardized to obtain the prediction compliance rate of the predicted output data sample set of new energy sources:

[0063]

[0064] Among them, X m L Typical scenario D for clustering to contribute to new energy prediction k The m-th new energy source predicted power output data sample set X m The predicted compliance rate of new energy output, where H is the total amount of samples taken over the time interval, h is the index of H, and x m,h For X m The new energy forecast compliance rate of the h-th new energy power output data sample. p X * n,h For the new energy power output data sample of the h-th sampling point during a certain period on day n, p Xn,h This is a sample of actual power output data of new energy sources at the h-th sampling point during a certain time period on day n.

[0065] Step 202: Set B% as the maximum allowable power outage risk rate for the wind-solar-storage combined system, and calculate the maximum power outage risk rate B% ​​in typical scenario D. k The quantile F of the predicted achievement rate in the ascending order of the new energy power output forecast data sample set. B％=max[floor(M×B%),1], according to quantile F B％ Index the data to find the Xm corresponding to the power outage risk rate B% ​​for each time period. L The backup safety inclusion coefficient P, which serves as a typical scenario for the clustering, is the predicted output of new energy sources. X *L (B%), by planning the proportion of renewable energy included in the reserve based on historical power output, makes the coefficient of renewable energy included in the reserve more realistic, where M is the sample set of the m-th renewable energy predicted power output data X. m The total number of samples in the data, where floor() is the floor function.

[0066] Step 3: Based on the predicted renewable energy output data and pumped storage reservoir capacity for the next time period, calculate the renewable energy reserve safety inclusion coefficient for the wind-solar-storage combined system based on pumped storage coordination. Solve for the reserve configuration capacity of the renewable energy output prediction cluster. Using the predicted renewable energy output value and pumped storage reservoir capacity for the next time period, quickly calculate the optimal renewable energy reserve inclusion ratio coefficient for the next time period, thereby determining the system's reserve capacity configuration; specifically, as follows... Figure 4 This includes the following sub-steps:

[0067] Step 301: Obtain the predicted output data of the wind-solar-storage combined system and the pumped storage reservoir capacity for the next time period;

[0068] Step 302: Calculate the similarity coefficient between the predicted renewable energy output data for the next time period and the cluster centers of typical scenarios in the predicted renewable energy output clustering. Select the typical scenario with the smallest similarity coefficient as the predicted renewable energy output scenario for the time period, and obtain the backup safety inclusion coefficient P of the predicted renewable energy output under the power supply failure risk rate B%. X *L (B%)

[0069] Step 303: Obtain the pumped storage status based on the pumped storage reservoir capacity for the next time period, and determine the pumped storage coordination coefficient P by combining it with the predicted maximum output of new energy sources for the next time period. C L :

[0070]

[0071] Among them, P C It is in pumped storage mode. q is the pumped storage power output coefficient. Q C Q represents the reservoir capacity of the pumped storage upper reservoir of the wind-solar-storage combined system. Cmin Q represents the minimum reservoir capacity of a pumped-storage hydroelectric system for a combined wind-solar-storage system. C^ To provide the reservoir capacity required for the pumped storage of the wind-solar-storage combined system to maintain full power generation in the next period, P Cmax P represents the maximum pumped storage output of the wind-solar-storage combined system. X * max This represents the maximum predicted output of new energy sources for the next time period.

[0072] Step 304: Using the pumped storage coordination coefficient P C L And the backup safety inclusion factor P of the predicted output of new energy sources under the power supply failure risk rate B%. X *L (B%), calculate the reserve safety inclusion factor P for the predicted output of new energy sources after coordination of pumped storage. X+C *L (B%):

[0073] P X+C *L (B%)=P X *L (B%)+P C L

[0074] Step 305: Based on the reserve safety inclusion factor P of the predicted output of new energy sources after pumped storage coordination. X+C *L (B%) and the minimum predicted output of new energy sources in the next period are used to obtain the reserve capacity S of the new energy output prediction cluster:

[0075] S = P X+C *L (B%)·P X * min .

[0076] This invention provides a method for configuring reserve capacity in wind-solar-storage integrated systems. First, based on an improved k-means clustering algorithm, it reduces the instability of clustering results caused by uncertainties in the initial cluster centers and the number of cluster scenarios. By generating typical scenarios of predicted renewable energy output for each time period based on historical data, it is more representative. Second, it can accurately calculate the renewable energy safety inclusion coefficient for each typical scenario under the constraint of power supply failure risk in the wind-solar-storage integrated system. Based on the predicted renewable energy output and pumped storage reservoir capacity for the next time period, it quickly determines the optimal proportion coefficient for renewable energy to be included in the reserve for the next time period, further reducing the minimum output value of thermal power units in the system for each time period and improving the system's renewable energy absorption capacity.

[0077] In another technical solution of the present invention, a computer-readable storage medium is also provided, storing a computer program that causes a computer to execute the described method for configuring standby capacity for a wind-solar-storage combined system.

[0078] In another technical solution of the present invention, an electronic device is also provided, including: a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the aforementioned method for configuring standby capacity for wind-solar-storage combined systems.

[0079] In the embodiments disclosed in this application, a computer storage medium may be a tangible medium that may contain or store programs for use by or in conjunction with an instruction execution system, apparatus, or device. The computer storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of computer storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0080] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed in this application can be implemented in electronic hardware or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0081] The above are merely preferred embodiments of the present invention. The scope of protection of the present invention is not limited to the above embodiments. All technical solutions falling within the scope of the present invention's concept are within the scope of protection of the present invention. It should be noted that for those skilled in the art, any improvements and modifications made without departing from the principle of the present invention should be considered within the scope of protection of the present invention.

Claims

1. A method for configuring standby capacity for wind-solar-storage integrated systems, characterized in that, Specifically, the steps include the following: Step 1: Based on the predicted output data of the wind-solar-storage combined system in a certain region, determine the optimal number of typical scenarios based on the improved K-means clustering algorithm, and generate typical scenarios for the predicted output clustering of new energy. Step 2: Using the calculation method of the renewable energy forecast output reserve security inclusion factor and the power supply loss risk rate constraint, determine the renewable energy forecast output reserve security inclusion factor for typical scenarios, including the following sub-steps: Step 201: Typical scenarios for clustering new energy predicted output D k The predicted output data sample set of new energy sources is combined with the corresponding actual output data sample set of new energy sources and standardized to obtain the prediction compliance rate of the predicted output data sample set of new energy sources: in, Typical scenarios for clustering to contribute to new energy forecasting D k The Middle m A sample set of predicted power output data for new energy sources X m The predicted rate of new energy power output meeting the target. k =1, 2, ..., K U , H The total number of samples taken at each time interval. h for H index, for X m The Middle h The new energy forecast compliance rate of a sample of new energy power output data. , p X n,h For the first n During a certain period of time on [date] h New energy predicted power output data samples from each sampling point p Xn,h For the first n During a certain period of time on [day] h Sample data of actual power output of new energy sources at each sampling point; Step 202, Setting B % represents the maximum permissible power outage risk rate for the wind-solar-storage combined system. Calculate the maximum power outage risk rate. B % in typical scenarios D k The quantiles corresponding to the predicted achievement rate in the ascending order of the new energy power output forecast data sample set. According to quantiles F B% Index the data to find the power outage risk rate for each time period. B % corresponds to Xm L As a typical scenario for the clustering, the backup safety inclusion coefficient for the predicted output of new energy sources is... ,in, M For the first m A sample set of predicted power output data for new energy sources X m Total number of samples in This is the floor function; Step 3: Based on the predicted renewable energy output data and pumped storage reservoir capacity for the next time period, calculate the renewable energy reserve safety inclusion coefficient of the wind-solar-storage combined system based on pumped storage coordination, and solve for the reserve configuration capacity of the renewable energy output prediction cluster, including the following sub-steps: Step 301: Obtain the predicted output data of the wind-solar-storage combined system and the pumped storage reservoir capacity for the next time period; Step 302: Calculate the similarity coefficient between the predicted renewable energy output data for the next time period and the cluster centers of typical scenarios in the predicted renewable energy output clustering. Select the typical scenario with the smallest similarity coefficient as the predicted renewable energy output scenario for the time period, and obtain the power supply failure risk rate. B The reserve safety inclusion factor for the predicted output of new energy sources is %. P X L ( B %); Step 303: Obtain the pumped storage status based on the pumped storage reservoir capacity for the next time period, and determine the pumped storage coordination coefficient by combining it with the predicted maximum output of new energy sources for the next time period. P C L : in, P C It is in pumped storage mode. , q This is the pumped storage power output coefficient. , The pumped storage capacity of the wind-solar-storage combined system is the upper reservoir capacity. Q Cmin This represents the minimum reservoir capacity for a pumped-storage hydroelectric system integrated with wind, solar, and energy storage. Q C ^ The reservoir capacity required for the pumped storage of the wind-solar-storage combined system to maintain full power generation in the next period. P Cmax This represents the maximum output of pumped storage in a wind-solar-storage combined system. P X max This represents the maximum predicted output of new energy sources for the next time period. Step 304: Using the pumped storage coordination coefficient P C L and power supply failure risk rate B The reserve safety inclusion factor for the predicted output of new energy sources is %. P X L ( B %), calculate the reserve safety inclusion factor for the predicted output of new energy sources after coordination of pumped storage. P X+C L ( B %): Step 305: Based on the reserve safety inclusion factor of the predicted output of new energy sources after pumped storage coordination. P X+C L ( B %) and the minimum predicted output of new energy sources in the next period. Obtain the backup capacity of new energy output prediction clustering S : 。 2.The method of claim 1, wherein, Step 1 includes the following sub-steps: Step 101: Collect actual power output data of new energy sources in a certain region at different times. P Xn ={ p Xn,h } and corresponding new energy power output forecast data P X n ={ p X n,h }, in, P Xn For the first n A sample set of actual new energy power output data for a certain period of time. p Xn,h For the first n During a certain period of time on [date] h Sample data of actual new energy output from each sampling point P X n For the first n The data sample set of predicted new energy power output for the corresponding time period of the day. p X n,h For the first n The corresponding time period of the day h Sample data of predicted new energy output from each sampling point; Step 102: Set the initial number of typical clustering scenarios and specify the cluster center for each typical clustering scenario; Step 103: Calculate the predicted power output data of new energy sources using the K-means clustering algorithm. P X n The similarity coefficients to each cluster center are used to select the new energy power output data with the lowest similarity coefficients. P X n They are categorized into the corresponding typical clustering scenarios; The predicted power output data of the new energy source P X n Similarity coefficients to each cluster center The calculation process is as follows: in, O j For the first j Cluster centers for a typical clustering scenario; Step 104: Calculate the mean of all new energy predicted output data samples in the same cluster typical scenario, and take the new energy predicted output data sample that is closest to the mean as the new cluster center of the cluster typical scenario; Step 105: If the new cluster centers of the typical clustering scenarios change, repeat the clustering process of steps 103-104 according to the new cluster centers; otherwise, determine whether the number of typical clustering scenarios is greater than 2. If so, decrement the number of clustering scenarios by 1, specify the cluster centers of each typical clustering scenario, and execute step 103; otherwise, calculate the average silhouette coefficient of all typical clustering scenarios based on the median. Z Select Z The maximum number of typical clustering scenarios is taken as the optimal value for the number of typical clustering scenarios, and typical scenarios for new energy power output prediction clustering are generated. 3.The method of claim 2, wherein, The median-based average profile coefficient Z The calculation process is as follows: in, u i This represents a sample of predicted power output data for new energy sources. P X i And all other new energy predicted power output data samples belonging to the typical clustering scenario D The median distance between them ; v i This represents a sample of predicted power output data for new energy sources. P X i And all new energy predicted power output data samples in the nearest neighbor clustering typical scenario E that do not belong to clustering typical scenario D The median of the distances between them. ; c For clustering scenarios D The number of sample data for the predicted output of new energy sources.

4. A computer readable storage medium storing a computer program, characterized in that, The computer program causes the computer to execute the backup capacity configuration method for wind-solar-storage combined systems as described in any one of claims 1-3.

5. An electronic device, comprising: include: The memory, the processor, and the computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, it implements the backup capacity configuration method for a wind-solar-storage combined system as described in any one of claims 1-3.

Images