An irradiance prediction method and device for distributed photovoltaics

Abstract

The present invention relates to the technical field of new energy power generation, and specifically provides an irradiance prediction method and device for distributed photovoltaic, including: clustering distributed photovoltaic sites in the regional irradiance numerical prediction grid point topology map, and obtaining a clustering square map corresponding to each clustering cluster; determining the irradiance numerical prediction values of each vertex of the clustering square map based on the vertex irradiance numerical prediction values of the grids to which the vertices of the clustering square map belong in the regional irradiance numerical prediction grid point topology map; determining the irradiance numerical prediction values of the distributed photovoltaic sites in each clustering cluster based on the irradiance numerical prediction values of each vertex of the clustering square map. The technical solution provided by the present invention has high calculation efficiency, good universality, is applicable to large-scale distributed photovoltaic clusters, can improve the prediction accuracy of distributed photovoltaic output, improve the fineness of the photovoltaic scenario, and has strong operability and popularization and application value.

Description

Technical Field

[0001] The present invention relates to the technical field of new energy power generation, and particularly to an irradiance prediction method and device for distributed photovoltaic power generation. Background Art

[0002] The short-term power prediction of new energy power generation refers to the power prediction for the next 1 to 3 days, which can provide decision-making support for the real-time scheduling of new energy and also provide a reference for the active power control of new energy power stations and the real-time trading of the power market. Distributed photovoltaic power generation has become an important direction for the development of new energy, with the characteristics of wide distribution and numerous sites. A large number of grid-connected distributed photovoltaic power stations have an impact on the safe and stable operation of the power grid. At present, the power prediction method for distributed photovoltaic clusters usually uses single-point numerical forecasting instead of regional numerical forecasting, and uses the irradiance-power conversion model to convert numerical forecasting into power. However, the accuracy of numerical forecasting is the main determinant of the prediction accuracy of distributed photovoltaic power generation. It is unreasonable to use single-point numerical forecasting instead of regional numerical forecasting, and there are certain errors in numerical forecasting itself.

[0003] There are many related studies on the power prediction of regional distributed photovoltaic power generation. Most of the current power prediction methods for distributed photovoltaic clusters take the prediction results of individual distributed photovoltaic power stations as the smallest unit, and calculate the prediction results of the region through the summation method or statistical upscaling method. The specific process is as follows: based on the longitude and latitude of the distributed photovoltaic sites, the numerical forecasts of each point are obtained; for each individual distributed photovoltaic station, a conversion model from irradiance to measured power is established; the numerical forecasts of each station are respectively converted into power; and the powers of each station are summed up as the regional prediction result.

[0004] The defects of the above solutions are as follows: 1) large data requirements: data such as measured power, measured irradiance, and numerical forecasting of each station are required; 2) low calculation efficiency and inapplicable to large-scale distributed photovoltaic clusters: it is necessary to establish models for each station separately, and the modeling time is long. As the cluster expands, the number of distributed photovoltaic power stations doubles, and the calculation efficiency will be greatly reduced.

[0005] At present, there is also a method for regional division. The specific operation method is as follows: the large region is divided into N small regions; models are established for each region separately: for each small region, the numerical forecasting of the central point is used instead of the regional numerical forecasting, and a conversion model between the numerical forecasting and the total power is established. The numerical forecasts of each small region are respectively converted into power; and then the powers of each small region are summed up as the power prediction result of the entire region.

[0006] The defects of the above - mentioned solution are as follows: 1) The method of dividing the area into grids is unreasonable: Since the distributed photovoltaic sites are not evenly distributed in the area, there will be a lot of invalid space in each small area after division; 2) The numerical prediction of the area represented by a single point is unreasonable: Different from the concentrated distribution of centralized photovoltaic fixed points, distributed photovoltaics show a dispersed cluster distribution, which is not concentrated and not uniform in the area, and there is a lot of invalid space in the small area, so the central point cannot represent the numerical prediction of the entire area;

[0007] 3) Completely relying on numerical prediction data: If there are obvious systematic errors in the numerical prediction data, it will have a greater impact on the accuracy. Summary of the Invention

[0008] In order to overcome the above - mentioned defects, the present invention proposes an irradiance prediction method and device for distributed photovoltaics.

[0009] In a first aspect, an irradiance prediction method for distributed photovoltaics is provided. The irradiance prediction method for distributed photovoltaics includes:

[0010] Cluster the distributed photovoltaic sites in the topological graph of the regional irradiance numerical prediction grid points, and obtain the clustering square graphs corresponding to each clustering cluster;

[0011] Based on the vertex irradiance numerical prediction values of the grids to which the vertices of the clustering square graph belong in the topological graph of the regional irradiance numerical prediction grid points, determine the irradiance numerical prediction values of the vertices of the clustering square graph;

[0012] Based on the irradiance numerical prediction values of the vertices of the clustering square graph, determine the irradiance numerical prediction values of the distributed photovoltaic sites in each clustering cluster.

[0013] Preferably, the clustering of the distributed photovoltaic sites in the topological graph of the regional irradiance numerical prediction grid points includes:

[0014] Take the meteorological monitoring device as the clustering center in the topological graph of the regional irradiance numerical prediction grid points, and cluster the distributed photovoltaic sites based on the distance between the clustering center and the distributed photovoltaic sites.

[0015] Furthermore, there is no intersection between the clustering square graphs corresponding to each clustering cluster. If there are distributed photovoltaic sites not covered by the clustering square graph, the distributed photovoltaic sites not covered by the clustering square graph are classified into the clustering square graph with the closest distance to it.

[0016] Preferably, the determination of the irradiance numerical prediction values of the vertices of the clustering square graph based on the vertex irradiance numerical prediction values of the grids to which the vertices of the clustering square graph belong in the topological graph of the regional irradiance numerical prediction grid points includes:

[0017] When the vertex of the clustering square graph is located at the vertex of the grid in the irradiance numerical prediction grid point topology map of its affiliated area, the irradiance numerical prediction value of the vertex of the clustering square graph is the irradiance numerical prediction value of the vertex of the grid in the irradiance numerical prediction grid point topology map of its affiliated area;

[0018] When the vertex of the clustering square graph is located on the boundary of the grid in the irradiance numerical prediction grid point topology map of its affiliated area, based on the irradiance numerical prediction values of the two endpoints of the boundary, the linear interpolation algorithm is used to determine the irradiance numerical prediction value of the vertex of the clustering square graph;

[0019] When the vertex of the clustering square graph is located inside the grid in the irradiance numerical prediction grid point topology map of its affiliated area, based on the irradiance numerical prediction values of the four vertices of the grid, the distance squared inverse ratio interpolation algorithm is used to determine the irradiance numerical prediction value of the vertex of the clustering square graph.

[0020] Furthermore, the calculation formula for determining the irradiance numerical prediction value of the vertex of the clustering square graph by using the linear interpolation algorithm is as follows:

[0021]

[0022] In the above formula, Z B is the irradiance numerical prediction value of vertex B of the clustering square graph, Za is the irradiance numerical prediction value of endpoint a of the boundary, Z b is the irradiance numerical prediction value of endpoint b of the boundary, d ba is the distance between vertex B of the clustering square graph and endpoint a of the boundary, d bb is the distance between vertex B of the clustering square graph and endpoint b of the boundary.

[0023] Furthermore, the calculation formula for determining the irradiance numerical prediction value of the vertex of the clustering square graph by using the distance squared inverse ratio interpolation algorithm is as follows:

[0024]

[0025] In the above formula, Z A is the irradiance numerical prediction value of vertex A of the clustering square graph, E i is the weight of vertex i of the grid in the irradiance numerical prediction grid point topology map of the area to which the vertex of the clustering square graph belongs, Z i is the irradiance numerical prediction value of vertex i of the grid in the irradiance numerical prediction grid point topology map of the area to which the vertex of the clustering square graph belongs.

[0026] Furthermore, the calculation formula for the weight of vertex i of the grid in the irradiance numerical prediction grid point topology map of the area to which the vertex of the clustering square graph belongs is as follows:

[0027]

[0028] In the above formula, d ai is the distance between vertex A of the clustering square diagram and vertex i of the grid in the topological diagram of the irradiance numerical prediction grid points in the area to which the vertex of the clustering square diagram belongs.

[0029] Preferably, the calculation formula for the irradiance numerical prediction value of the distributed photovoltaic sites in each clustering cluster is as follows:

[0030]

[0031] In the above formula, Z is the irradiance numerical prediction value of the distributed photovoltaic sites in the clustering cluster, and Z j is the irradiance numerical prediction value of vertex j of the clustering square diagram corresponding to the clustering cluster, and Q j is the weight of vertex j of the clustering square diagram corresponding to the clustering cluster.

[0032] Furthermore, the process of obtaining the weight of vertex j of the clustering square diagram corresponding to the clustering cluster includes:

[0033] Solving the pre-constructed maximum correlation model to obtain the weight of vertex j of the clustering square diagram corresponding to the clustering cluster.

[0034] Furthermore, the pre-constructed maximum correlation model includes: an objective function with the goal of maximizing the correlation coefficients between the irradiance numerical prediction values of the distributed photovoltaic sites in the clustering cluster and the measured total power and the meteorological monitoring irradiance respectively, and constraint conditions configured for the irradiance prediction of distributed photovoltaics.

[0035] Furthermore, the calculation formula of the objective function is as follows:

[0036] max(r ZP )

[0037] max(r ZR )

[0038] In the above formula, r ZP is the correlation coefficient between the irradiance numerical prediction value of the distributed photovoltaic sites in the clustering cluster and the measured total power, and r ZR is the correlation coefficient between the irradiance numerical prediction value of the distributed photovoltaic sites in the clustering cluster and the meteorological monitoring irradiance.

[0039] Furthermore, the calculation formula of the correlation coefficient between the irradiance numerical prediction value of the distributed photovoltaic sites in the clustering cluster and the measured total power is as follows:

[0040]

[0041] The calculation formula for the correlation coefficient between the predicted irradiance value of the distributed PV sites in the clustering cluster and the meteorological monitored irradiance is as follows:

[0042]

[0043] In the above formula, Z is the predicted irradiance value of the distributed PV sites in the clustering cluster, P is the measured total power, R is the meteorological monitored irradiance, E is the mathematical expectation function, D is the variance function, μ1 is the mean value of the predicted irradiance value of the distributed PV sites in the clustering cluster, μ2 is the mean value of the measured total power, and μ3 is the mean value of the meteorological monitored irradiance.

[0044] Furthermore, the mathematical model of the constraint condition is:

[0045]

[0046] Q A +Q B +Q C +Q D =1

[0047] In the above formula, Q A 、Q B 、Q C 、Q D are the weights of the four vertices of the clustering square corresponding to the clustering cluster respectively.

[0048] On the second aspect, an irradiance prediction device for distributed PV is provided. The irradiance prediction device for distributed PV includes:

[0049] A clustering module, configured to cluster the distributed PV sites in the regional irradiance numerical prediction grid point topology map and obtain the clustering square corresponding to each clustering cluster;

[0050] A first determination module, configured to determine the predicted irradiance value of each vertex of the clustering square based on the predicted irradiance value of the vertex of the grid where each vertex of the clustering square belongs in the regional irradiance numerical prediction grid point topology map;

[0051] A second determination module, configured to determine the predicted irradiance value of the distributed PV sites in each clustering cluster based on the predicted irradiance value of each vertex of the clustering square.

[0052] Preferably, the clustering module is specifically configured to:

[0053] Take the meteorological monitoring device as the clustering center in the regional irradiance numerical prediction grid point topology map, and cluster the distributed PV sites based on the distance between the clustering center and the distributed PV sites.

[0054] Further, there is no intersection between the clustering square diagrams corresponding to the respective clustering clusters. If there is a distributed PV site not covered by the clustering square diagram, the distributed PV site not covered by the clustering square diagram is classified into the clustering square diagram closest to it.

[0055] Preferably, the first determination module is specifically configured to:

[0056] When the vertex of the clustering square diagram is located at the vertex of the grid in the irradiance numerical prediction grid point topology map of its affiliated area, the irradiance numerical prediction value of the vertex of the clustering square diagram is the irradiance numerical prediction value of the vertex of the grid in the irradiance numerical prediction grid point topology map of its affiliated area;

[0057] When the vertex of the clustering square diagram is located on the boundary of the grid in the irradiance numerical prediction grid point topology map of its affiliated area, based on the irradiance numerical prediction values of the two endpoints of the boundary, a linear interpolation algorithm is used to determine the irradiance numerical prediction value of the vertex of the clustering square diagram;

[0058] When the vertex of the clustering square diagram is located inside the grid in the irradiance numerical prediction grid point topology map of its affiliated area, based on the irradiance numerical prediction values of the four vertices of the grid, a distance squared inverse ratio interpolation algorithm is used to determine the irradiance numerical prediction value of the vertex of the clustering square diagram.

[0059] Further, the calculation formula for determining the irradiance numerical prediction value of the vertex of the clustering square diagram by using the linear interpolation algorithm is as follows:

[0060]

[0061] In the above formula, Z B is the irradiance numerical prediction value of vertex B of the clustering square diagram, Za is the irradiance numerical prediction value of endpoint a of the boundary, Z b is the irradiance numerical prediction value of endpoint b of the boundary, d ba is the distance between vertex B of the clustering square diagram and endpoint a of the boundary, d bb is the distance between vertex B of the clustering square diagram and endpoint b of the boundary.

[0062] Further, the calculation formula for determining the irradiance numerical prediction value of the vertex of the clustering square diagram by using the distance squared inverse ratio interpolation algorithm is as follows:

[0063]

[0064] In the above formula, Z A is the irradiance numerical prediction value of vertex A of the clustering square diagram, E iis the weight of vertex i of the grid in the irradiance numerical prediction grid point topology map of the area to which the vertex of the clustering square graph belongs, Z i is the irradiance numerical prediction value of vertex i of the grid in the irradiance numerical prediction grid point topology map of the area to which the vertex of the clustering square graph belongs.

[0065] Furthermore, the calculation formula for the weight of vertex i of the grid in the irradiance numerical prediction grid point topology map of the area to which the vertex of the clustering square graph belongs is as follows:

[0066]

[0067] In the above formula, d ai is the distance between vertex A of the clustering square graph and vertex i of the grid in the irradiance numerical prediction grid point topology map of the area to which the vertex of the clustering square graph belongs.

[0068] Preferably, the calculation formula for the irradiance numerical prediction value of the distributed photovoltaic sites in each clustering cluster is as follows:

[0069]

[0070] In the above formula, Z is the irradiance numerical prediction value of the distributed photovoltaic sites in the clustering cluster, Z j is the irradiance numerical prediction value of vertex j of the clustering square graph corresponding to the clustering cluster, Q j is the weight of vertex j of the clustering square graph corresponding to the clustering cluster.

[0071] Furthermore, the process of obtaining the weight of vertex j of the clustering square graph corresponding to the clustering cluster includes:

[0072] Solving the pre-constructed maximum correlation model to obtain the weight of vertex j of the clustering square graph corresponding to the clustering cluster.

[0073] Furthermore, the pre-constructed maximum correlation model includes: an objective function with the maximum correlation coefficient between the irradiance numerical prediction value of the distributed photovoltaic sites in the clustering cluster and the measured total power and the meteorological monitoring irradiance respectively as the goal, and constraint conditions configured for the irradiance prediction of the distributed photovoltaic.

[0074] Furthermore, the calculation formula for the objective function is as follows:

[0075] max(r ZP )

[0076] max(r ZR )

[0077] In the above formula, r ZP is the correlation coefficient between the irradiance numerical prediction value of the distributed photovoltaic sites in the clustering cluster and the measured total power, rZR The correlation coefficient between the predicted irradiance value of the distributed PV sites in the clustering cluster and the meteorological monitored irradiance.

[0078] Furthermore, the calculation formula of the correlation coefficient between the predicted irradiance value of the distributed PV sites in the clustering cluster and the measured total power is as follows:

[0079]

[0080] The calculation formula of the correlation coefficient between the predicted irradiance value of the distributed PV sites in the clustering cluster and the meteorological monitored irradiance is as follows:

[0081]

[0082] In the above formula, Z is the predicted irradiance value of the distributed PV sites in the clustering cluster, P is the measured total power, R is the meteorological monitored irradiance, E is the mathematical expectation function, D is the variance function, μ1 is the mean value of the predicted irradiance value of the distributed PV sites in the clustering cluster, μ2 is the mean value of the measured total power, and μ3 is the mean value of the meteorological monitored irradiance.

[0083] Furthermore, the mathematical model of the constraint condition is:

[0084]

[0085] Q A +Q B +Q C +Q D =1

[0086] In the above formula, Q A 、Q B 、Q C 、Q D are the weights of the four vertices of the clustering square diagram corresponding to the clustering cluster respectively.

[0087] In the third aspect, a computer device is provided, including: one or more processors;

[0088] The processor is used to store one or more programs;

[0089] When the one or more programs are executed by the one or more processors, the irradiance prediction method for distributed PV as described above is implemented.

[0090] In the fourth aspect, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed, the irradiance prediction method for distributed PV as described above is implemented.

[0091] One or more of the above technical solutions of the present invention have at least one or more of the following beneficial effects:

[0092] The present invention provides an irradiance prediction method and device for distributed photovoltaic, including: clustering distributed photovoltaic sites in the topological map of the regional irradiance numerical prediction grid points, and obtaining the corresponding clustering square maps for each clustering cluster; based on the vertex irradiance numerical prediction values of the grids to which the vertices of the clustering square map belong in the topological map of the regional irradiance numerical prediction grid points, determining the irradiance numerical prediction values of the vertices of the clustering square map; and determining the irradiance numerical prediction values of the distributed photovoltaic sites in each clustering cluster based on the irradiance numerical prediction values of the vertices of the clustering square map. The technical solution provided by the present invention makes full use of the topological relationship among the meteorological monitoring device, the distributed photovoltaic site, and the numerical prediction grid points, clusters the distributed photovoltaic stations with the meteorological monitoring device as the center, forms clustering square maps with no intersections and high site density, has less invalid space in the divided clustering square maps, and the division method is more reasonable;

[0093] Further, for the technical solution provided by the present invention, based on the vertex irradiance numerical prediction values of the grids to which the vertices of the clustering square map belong in the topological map of the regional irradiance numerical prediction grid points, determining the irradiance numerical prediction values of the vertices of the clustering square map, and using the irradiance numerical prediction values of the four vertices to correct the irradiance numerical prediction values of the vertices of the clustering square map is more reasonable than using a single-point numerical prediction to replace the region;

[0094] Furthermore, the technical solution provided by the present invention does not completely rely on numerical prediction, and will correct future numerical predictions according to historical monitored meteorology in real time, making the numerical prediction data more accurate and improving the prediction accuracy;

[0095] In summary, the technical solution provided by the present invention makes full use of the topological relationship among the meteorological monitoring device, the distributed photovoltaic site, and the numerical prediction grid, corrects the numerical prediction data in real time, has a small data demand and high calculation efficiency, has good universality, is applicable to large-scale distributed photovoltaic clusters, can improve the prediction accuracy of distributed photovoltaic output, improve the refinement degree of the photovoltaic scenario, and has strong operability and popularization and application value. It can provide decision support for real-time dispatching of new energy, and also provide reference for active power control of new energy power stations and real-time trading in the power market. Description of the Drawings

[0096] Figure 1 is a schematic flow chart of the main steps of the irradiance prediction method for distributed photovoltaic according to an embodiment of the present invention;

[0097] Figure 2 is a schematic diagram of the topological relationship between the meteorological monitoring device and the distributed photovoltaic site according to an embodiment of the present invention;

[0098] Figure 3 Schematic diagram of the photovoltaic site of the embodiment of the present invention on the grid data node;

[0099] Figure 4 Schematic diagram of the photovoltaic site of the embodiment of the present invention on the grid data boundary;

[0100] Figure 5 Schematic diagram of the photovoltaic site of the embodiment of the present invention within the grid of the grid data;

[0101] Figure 6 Comparison chart of sunny day prediction effects of the embodiment of the present invention;

[0102] Figure 7 Comparison chart of cloudy and rainy day prediction effects of the embodiment of the present invention. Detailed implementation manners

[0103] The following further elaborates on the detailed implementation manners of the present invention with reference to the accompanying drawings.

[0104] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the scope of protection of the present invention.

[0105] Embodiment 1

[0106] As introduced in the background art, in recent years, distributed photovoltaics has become an important direction for the development of new energy. Distributed photovoltaics has the characteristics of numerous points and wide areas, and its randomness will significantly affect the economy and stability of the distribution network. Currently, the distributed photovoltaic cluster power prediction method usually uses single-point numerical forecasting instead of regional numerical forecasting, and uses the irradiance-power conversion model to convert numerical forecasting into power. However, the accuracy of numerical forecasting is the main determinant of the distributed photovoltaic prediction accuracy. It is unreasonable to replace the region with single-point numerical forecasting, and there are certain errors in numerical forecasting itself.

[0107] To improve the above problems, the present invention uses the clustering and optimization weighting method to fully utilize the topological relationship among the grid-based numerical forecasting, meteorological monitoring devices, and distributed photovoltaic sites, converts the stochastic optimization problem into a deterministic optimization problem under multiple scenarios, and more reasonably applies the numerical forecasting data to the distributed photovoltaic cluster to improve the refinement level of the distributed photovoltaic scenario. At the same time, based on the meteorological monitoring data, the future irradiance forecast is corrected to improve the accuracy of power prediction. And this method is simple to operate and applicable to regional distributed photovoltaic clusters. Refer to the attached Figure 1, Figure 1 is a schematic diagram of the main steps of the irradiance prediction method for distributed photovoltaic systems according to an embodiment of the present invention. As shown in Figure 1 shown, the irradiance prediction method for distributed photovoltaic systems in the embodiments of the present invention mainly includes the following steps:

[0108] Step S101: Cluster the distributed photovoltaic sites in the regional irradiance numerical prediction grid point topology map, and obtain the corresponding clustering square diagrams for each clustering cluster;

[0109] Step S102: Based on the vertex irradiance numerical prediction values of the grids to which the vertices of the clustering square diagram belong in the regional irradiance numerical prediction grid point topology map, determine the irradiance numerical prediction values of the vertices of the clustering square diagram;

[0110] Step S103: Based on the irradiance numerical prediction values of the vertices of the clustering square diagram, determine the irradiance numerical prediction values of the distributed photovoltaic sites in each clustering cluster.

[0111] The present invention provides an irradiance prediction method and device for distributed photovoltaic systems, including: clustering the distributed photovoltaic sites in the regional irradiance numerical prediction grid point topology map, and obtaining the corresponding clustering square diagrams for each clustering cluster; based on the vertex irradiance numerical prediction values of the grids to which the vertices of the clustering square diagram belong in the regional irradiance numerical prediction grid point topology map, determine the irradiance numerical prediction values of the vertices of the clustering square diagram; based on the irradiance numerical prediction values of the vertices of the clustering square diagram, determine the irradiance numerical prediction values of the distributed photovoltaic sites in each clustering cluster. The technical solution provided by the present invention makes full use of the topological relationship among the meteorological monitoring device, the distributed photovoltaic site, and the numerical prediction grid points, clusters the distributed photovoltaic stations with the meteorological monitoring device as the center, forms a clustering square diagram with no intersection and high site density, and the invalid space of the divided clustering square diagram is less, and the division method is more reasonable;

[0112] Furthermore, for the vertex irradiance numerical prediction values of the grids to which the vertices of the clustering square diagram belong in the regional irradiance numerical prediction grid point topology map, the technical solution provided by the present invention determines the irradiance numerical prediction values of the vertices of the clustering square diagram, and corrects the irradiance numerical prediction values of the vertices of the clustering square diagram by using the irradiance numerical prediction values of the four vertices, which is more reasonable than using a single-point numerical prediction to replace the region;

[0113] Still further, the technical solution provided by the present invention does not completely rely on numerical prediction, but will revise the future numerical prediction according to historical monitored meteorology in real time, making the numerical prediction data more accurate and improving the prediction accuracy;

[0114] In summary, the technical solution provided by the present invention makes full use of the topological relationship among the meteorological monitoring device, the distributed photovoltaic site, and the numerical prediction grid, corrects the numerical prediction data in real time, has a high data demand and calculation efficiency, has good universality, is applicable to large-scale distributed photovoltaic clusters, can improve the prediction accuracy of distributed photovoltaic output, improve the refinement degree of photovoltaic scenarios, and has strong operability and popularization and application value. It can provide decision support for the real-time dispatching of new energy, and also provide references for the active power control of new energy power stations and the real-time trading of the power market. The above solution will be elaborated in detail below.

[0115] Embodiment 2

[0116] In this embodiment, before clustering the distributed photovoltaic sites in the topological map of the regional irradiance numerical prediction grid points, it is first necessary to obtain data and establish a topological relationship map among the meteorological monitoring device, the distributed photovoltaic site, and the numerical prediction grid points, which can be obtained in the following manner:

[0117] Respectively obtain the geographical locations (latitude and longitude), topographies and landforms, and topological relationship maps of the distributed photovoltaic sites and the locations of the meteorological monitoring devices, and the actual and planned installed capacities of each distributed photovoltaic power station, and establish a topological relationship map among the meteorological monitoring device, the distributed photovoltaic site, and the numerical prediction grid points;

[0118] Produce grid-based regional numerical weather forecast data, including total solar radiation, direct radiation, diffuse radiation, wind speed, wind direction, temperature, cumulative precipitation, air pressure, and humidity. Obtain meteorological monitoring data and measured power monitoring data of each distributed site.

[0119] Furthermore, the clustering of the distributed photovoltaic sites in the topological map of the regional irradiance numerical prediction grid points includes:

[0120] In the topological map of the regional irradiance numerical prediction grid points, using the meteorological monitoring device as the clustering center, cluster the distributed photovoltaic sites based on the distance between the clustering center and the distributed photovoltaic sites.

[0121] In one embodiment, the drawing of the clustering square diagram corresponding to each clustering cluster can be carried out in the following manner:

[0122] Taking the meteorological monitoring device as the center, through clustering, map the distributed photovoltaic sites in the region to each meteorological monitoring site in a one-to-many mapping relationship, and divide the region into N meteorological monitoring circles with rectangles, as Figure 2As shown. And each clustered square diagram has the following characteristics: it contains as many distributed sites as possible, adjacent rectangles do not cross, and the site density in the rectangles is as large as possible. If there are individual remaining distributed sites, consider including them in the nearest meteorological monitoring circle. However, due to the high-density distributed photovoltaic power stations within the circle, the corresponding numerical prediction data is still extracted based on the original meteorological monitoring circle as the reference area.

[0123] Among them, there is no intersection between the clustered square diagrams corresponding to each clustering cluster. If there are distributed photovoltaic sites not covered by the clustered square diagrams, then the distributed photovoltaic sites not covered by the clustered square diagrams are included in the clustered square diagram with the closest distance to them.

[0124] In this embodiment, the numerical prediction data at each vertex of each clustered square is obtained through a non-linear interpolation method. Specifically, determining the irradiance numerical prediction value at each vertex of the clustered square diagram based on the vertex irradiance numerical prediction values of the grids to which the vertices of the clustered square diagram belong in the regional irradiance numerical prediction grid point topology diagram includes: Figure 4 When the vertex of the clustered square diagram is located at the vertex of the grid in the regional irradiance numerical prediction grid point topology diagram to which it belongs, the irradiance numerical prediction value of the vertex of the clustered square diagram is the irradiance numerical prediction value of the vertex of the grid in the regional irradiance numerical prediction grid point topology diagram to which it belongs. For example, as

[0125] shown, point C in the figure is located at the vertex of the grid in the regional irradiance numerical prediction grid point topology diagram to which it belongs; Figure 3 As shown, point C in the figure is located at the vertex of the grid in the regional irradiance numerical prediction grid point topology diagram to which it belongs;

[0126] See Figure 4 , when the vertex of the clustered square diagram is located on the boundary of the grid in the regional irradiance numerical prediction grid point topology diagram to which it belongs, based on the irradiance numerical prediction values of the two endpoints of the boundary, a linear interpolation algorithm is used to determine the irradiance numerical prediction value of the vertex of the clustered square diagram;

[0127] See Figure 5 , when the vertex of the clustered square diagram is located inside the grid in the regional irradiance numerical prediction grid point topology diagram to which it belongs, based on the irradiance numerical prediction values of the four vertices of the grid, an inverse distance squared interpolation algorithm is used to determine the irradiance numerical prediction value of the vertex of the clustered square diagram.

[0128] In one embodiment, the calculation formula for determining the irradiance numerical prediction value of the vertex of the clustered square diagram using the linear interpolation algorithm is as follows:

[0129]

[0130] In the above formula, Z Bis the predicted irradiance value of vertex B of the clustering square diagram, Za is the predicted irradiance value of endpoint a of the boundary, Z b is the predicted irradiance value of endpoint b of the boundary, d ba is the distance between vertex B of the clustering square diagram and endpoint a of the boundary, d bb is the distance between vertex B of the clustering square diagram and endpoint b of the boundary.

[0131] In one embodiment, the calculation formula for determining the predicted irradiance value of the vertex of the clustering square diagram using the inverse square distance interpolation algorithm is as follows:

[0132]

[0133] In the above formula, Z A is the predicted irradiance value of vertex A of the clustering square diagram, E i is the weight of vertex i of the grid in the irradiance value prediction grid point topology diagram of the area to which the vertex of the clustering square diagram belongs, Z i is the predicted irradiance value of vertex i of the grid in the irradiance value prediction grid point topology diagram of the area to which the vertex of the clustering square diagram belongs.

[0134] In one embodiment, the calculation formula for the weight of vertex i of the grid in the irradiance value prediction grid point topology diagram of the area to which the vertex of the clustering square diagram belongs is as follows:

[0135]

[0136] In the above formula, d ai is the distance between vertex A of the clustering square diagram and vertex i of the grid in the irradiance value prediction grid point topology diagram of the area to which the vertex of the clustering square diagram belongs.

[0137] In this embodiment, the calculation formula for the predicted irradiance value of the distributed photovoltaic sites in each clustering cluster is as follows:

[0138]

[0139] In the above formula, Z is the predicted irradiance value of the distributed photovoltaic sites in the clustering cluster, Z j is the predicted irradiance value of vertex j of the clustering square diagram corresponding to the clustering cluster, Q j is the weight of vertex j of the clustering square diagram corresponding to the clustering cluster.

[0140] Further, the determination of the weight coefficient is an optimization weighting problem of a bi-objective function. The two objective functions are respectively to maximize the cross-correlation coefficients between the numerical prediction irradiance within the square graph and the measured total power and the meteorological monitoring irradiance. The cross-correlation coefficient characterizes the spatial correlation between multiple time series. Generally, if the cross-correlation coefficient is greater than 0.7, the two groups of sequences have a strong cross-correlation, indicating that their changes are very synchronous. It is known that there is a strong correlation between the measured irradiance and the measured power. Therefore, the numerical prediction irradiance with better correlations with both the measured irradiance and the measured total power within the square graph can better represent the numerical prediction irradiance of distributed photovoltaics in a small area. The process for obtaining the weight of vertex j of the clustering square graph corresponding to the clustering cluster includes:

[0141] Solving the pre-constructed maximum correlation model to obtain the weight of vertex j of the clustering square graph corresponding to the clustering cluster.

[0142] In one embodiment, the pre-constructed maximum correlation model includes: an objective function aiming to maximize the correlation coefficients between the numerical prediction values of the irradiance of distributed photovoltaic sites in the clustering cluster and the measured total power and the meteorological monitoring irradiance respectively, and constraint conditions configured for the irradiance prediction of distributed photovoltaics.

[0143] In one embodiment, the calculation formula of the objective function is as follows:

[0144] max(r ZP )

[0145] max(r ZR )

[0146] In the above formula, r ZP is the correlation coefficient between the numerical prediction value of the irradiance of distributed photovoltaic sites in the clustering cluster and the measured total power, and r ZR is the correlation coefficient between the numerical prediction value of the irradiance of distributed photovoltaic sites in the clustering cluster and the meteorological monitoring irradiance.

[0147] Further, the calculation formula of the correlation coefficient between the numerical prediction value of the irradiance of distributed photovoltaic sites in the clustering cluster and the measured total power is as follows:

[0148]

[0149] The calculation formula of the correlation coefficient between the numerical prediction value of the irradiance of distributed photovoltaic sites in the clustering cluster and the meteorological monitoring irradiance is as follows:

[0150]

[0151] In the above formula, Z is the numerical prediction value of the irradiance of distributed photovoltaic sites in the clustering cluster, P is the measured total power, R is the meteorological monitoring irradiance, E is the mathematical expectation function, D is the variance function, μ1 is the mean value of the numerical prediction value of the irradiance of distributed photovoltaic sites in the clustering cluster, μ2 is the mean value of the measured total power, and μ3 is the mean value of the meteorological monitoring irradiance.

[0152] Furthermore, the mathematical model of the constraint condition is:

[0153]

[0154] Q A +Q B +Q C +Q D = 1

[0155] In the above formula, Q A , Q B , Q C , Q D are the weights of the four vertices of the clustering square diagram corresponding to the clustering cluster respectively.

[0156] In this embodiment, an intelligent algorithm is used to correct the numerical prediction irradiance with the meteorological monitoring irradiance, a model is established with the historical meteorological monitoring irradiance and the historical numerical prediction irradiance, the error law characteristics of the numerical prediction data are found, and the future numerical prediction irradiance is corrected.

[0157] The following takes a distributed photovoltaic power station in a certain area for verification. For short-term prediction correction, the predicted irradiance data for the next day is iteratively corrected, and then the irradiance-power conversion model established with the historical measured irradiance and measured power is used to convert the original numerical prediction and the corrected numerical prediction into power. The prediction effect on sunny days is as Figure 6 shown, and the prediction effect on cloudy and rainy days is as Figure 7 shown.

[0158] The calculation of each error index is shown in Table 1:

[0159] Table 1

[0160]

[0161] The calculation example results show that each error index of the ultra-short-term predicted power is better than that of the short-term predicted power. Applying the method of the present invention can effectively improve the ultra-short-term power prediction accuracy of new energy power generation.

[0162] Embodiment 3

[0163] Based on the same inventive concept, the present invention also provides an irradiance prediction device for distributed photovoltaics, and the irradiance prediction device for distributed photovoltaics includes:

[0164] A clustering module, which is used to cluster distributed photovoltaic sites in the topological map of grid points of regional irradiance numerical prediction, and obtain a clustering square map corresponding to each clustering cluster;

[0165] A first determination module, which is used to determine the irradiance numerical prediction values of the vertices of the clustering square map based on the irradiance numerical prediction values of the vertices of the grid to which the vertices of the clustering square map belong in the topological map of grid points of regional irradiance numerical prediction;

[0166] A second determination module, which is used to determine the irradiance numerical prediction values of the distributed photovoltaic sites in each clustering cluster based on the irradiance numerical prediction values of the vertices of the clustering square map.

[0167] Preferably, the clustering module is specifically used for:

[0168] Taking a meteorological monitoring device as a clustering center in the topological map of grid points of regional irradiance numerical prediction, and clustering the distributed photovoltaic sites based on the distance between the clustering center and the distributed photovoltaic sites.

[0169] Furthermore, there is no intersection between the clustering square maps corresponding to each clustering cluster. If there are distributed photovoltaic sites not covered by the clustering square map, then the distributed photovoltaic sites not covered by the clustering square map are classified into the clustering square map closest to them.

[0170] Preferably, the first determination module is specifically used for:

[0171] When the vertex of the clustering square map is located at the vertex of the grid in the topological map of grid points of regional irradiance numerical prediction to which it belongs, the irradiance numerical prediction value of the vertex of the clustering square map is the irradiance numerical prediction value of the vertex of the grid in the topological map of grid points of regional irradiance numerical prediction to which it belongs;

[0172] When the vertex of the clustering square map is located on the boundary of the grid in the topological map of grid points of regional irradiance numerical prediction to which it belongs, based on the irradiance numerical prediction values of the two endpoints of the boundary, a linear interpolation algorithm is used to determine the irradiance numerical prediction value of the vertex of the clustering square map;

[0173] When the vertex of the clustering square map is located inside the grid in the topological map of grid points of regional irradiance numerical prediction to which it belongs, based on the irradiance numerical prediction values of the four vertices of the grid, a distance squared inverse interpolation algorithm is used to determine the irradiance numerical prediction value of the vertex of the clustering square map.

[0174] Furthermore, the calculation formula for determining the irradiance numerical prediction value of the vertex of the clustering square map by using the linear interpolation algorithm is as follows:

[0175]

[0176] In the above formula, ZB is the predicted irradiance value at vertex B of the clustering square diagram, Za is the predicted irradiance value at endpoint a of the boundary, and Z b is the predicted irradiance value at endpoint b of the boundary, and d ba is the distance between vertex B of the clustering square diagram and endpoint a of the boundary, and d bb is the distance between vertex B of the clustering square diagram and endpoint b of the boundary.

[0177] Furthermore, the calculation formula for determining the predicted irradiance value at the vertex of the clustering square diagram using the inverse square distance interpolation algorithm is as follows:

[0178]

[0179] In the above formula, Z A is the predicted irradiance value at vertex A of the clustering square diagram, and E i is the weight of vertex i of the grid in the topological diagram of the predicted irradiance value grid points in the area where the vertex of the clustering square diagram is located, and Z i is the predicted irradiance value at vertex i of the grid in the topological diagram of the predicted irradiance value grid points in the area where the vertex of the clustering square diagram is located.

[0180] Furthermore, the calculation formula for the weight of vertex i of the grid in the topological diagram of the predicted irradiance value grid points in the area where the vertex of the clustering square diagram is located is as follows:

[0181]

[0182] In the above formula, d ai is the distance between vertex A of the clustering square diagram and vertex i of the grid in the topological diagram of the predicted irradiance value grid points in the area where the vertex of the clustering square diagram is located.

[0183] Preferably, the calculation formula for the predicted irradiance value of the distributed photovoltaic sites in each clustering cluster is as follows:

[0184]

[0185] In the above formula, Z is the predicted irradiance value of the distributed photovoltaic sites in the clustering cluster, and Z j is the predicted irradiance value at vertex j of the clustering square diagram corresponding to the clustering cluster, and Q j is the weight of vertex j of the clustering square diagram corresponding to the clustering cluster.

[0186] Furthermore, the process of obtaining the weight of vertex j of the clustering square diagram corresponding to the clustering cluster includes:

[0187] Solve the pre-constructed maximum correlation model to obtain the weight of vertex j of the clustering square graph corresponding to the clustering cluster.

[0188] Furthermore, the pre-constructed maximum correlation model includes: an objective function with the goal of maximizing the correlation coefficients between the predicted irradiance values of distributed photovoltaic sites in the clustering cluster and the measured total power and the meteorological monitoring irradiance respectively, and constraint conditions configured for the irradiance prediction of distributed photovoltaics.

[0189] Furthermore, the calculation formula of the objective function is as follows:

[0190] max(r ZP )

[0191] max(r ZR )

[0192] In the above formula, r ZP is the correlation coefficient between the predicted irradiance value of the distributed photovoltaic site in the clustering cluster and the measured total power, and r ZR is the correlation coefficient between the predicted irradiance value of the distributed photovoltaic site in the clustering cluster and the meteorological monitoring irradiance.

[0193] Furthermore, the calculation formula of the correlation coefficient between the predicted irradiance value of the distributed photovoltaic site in the clustering cluster and the measured total power is as follows:

[0194]

[0195] The calculation formula of the correlation coefficient between the predicted irradiance value of the distributed photovoltaic site in the clustering cluster and the meteorological monitoring irradiance is as follows:

[0196]

[0197] In the above formula, Z is the predicted irradiance value of the distributed photovoltaic site in the clustering cluster, P is the measured total power, R is the meteorological monitoring irradiance, E is the mathematical expectation function, D is the variance function, μ1 is the mean value of the predicted irradiance value of the distributed photovoltaic site in the clustering cluster, μ2 is the mean value of the measured total power, and μ3 is the mean value of the meteorological monitoring irradiance.

[0198] Furthermore, the mathematical model of the constraint condition is:

[0199]

[0200] Q A +Q B +Q C +Q D =1

[0201] In the above formula, Q A, Q B , Q C , Q D are the weights of the four vertices of the clustering square diagram corresponding to the clustering cluster respectively.

[0202] Embodiment 4

[0203] Based on the same inventive concept, the present invention also provides a computer device, which includes a processor and a memory. The memory is used to store a computer program, and the computer program includes program instructions. The processor is used to execute the program instructions stored in the computer storage medium. The processor may be a Central Processing Unit (CPU), or may also be other general-purpose processors, Digital Signal Processors (DSPs), Application Specific Integrated Circuits (ASICs), Field-Programmable Gate Arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. It is the computing core and control core of the terminal, and is suitable for implementing one or more instructions. Specifically, it is suitable for loading and executing one or more instructions in the computer storage medium to implement the corresponding method flow or corresponding function, so as to implement the steps of a method for irradiance prediction for distributed photovoltaic in the above embodiment.

[0204] Embodiment 5

[0205] Based on the same inventive concept, the present invention also provides a storage medium, specifically a computer-readable storage medium (Memory). The computer-readable storage medium is a memory device in a computer device and is used to store programs and data. It can be understood that the computer-readable storage medium here can include both the built-in storage medium in the computer device and, of course, the extended storage medium supported by the computer device. The computer-readable storage medium provides a storage space, and this storage space stores the operating system of the terminal. And, one or more instructions suitable for being loaded and executed by the processor are also stored in this storage space. These instructions may be one or more computer programs (including program codes). It should be noted that the computer-readable storage medium here can be a high-speed RAM memory or a non-volatile memory, such as at least one disk memory. One or more instructions stored in the computer-readable storage medium can be loaded and executed by the processor to implement the steps of a method for irradiance prediction for distributed photovoltaic in the above embodiment.

[0206] Those skilled in the art should understand that the embodiments of the present invention can be provided as a method, a system, or a computer program product. Therefore, the present invention can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0207] The present invention is described with reference to the flowcharts and / or block diagrams of methods, apparatuses (systems), and computer program products according to the embodiments of the present invention. It should be understood that each flow and / or block in the flowchart and / or block diagram, as well as the combination of flows and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, such that the instructions executed by the processor of the computer or other programmable data processing devices generate means for implementing the functions specified in Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0208] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing devices to work in a specific manner, such that the instructions stored in the computer-readable memory generate a manufactured article including instruction means that implement the functions specified in Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0209] These computer program instructions can also be loaded onto a computer or other programmable data processing devices, such that a series of operation steps are executed on the computer or other programmable devices to generate a computer-implemented process. Thus, the instructions executed on the computer or other programmable devices provide steps for implementing the functions specified in Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0210] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the above embodiments, those of ordinary skill in the art should understand that: still can modify the specific implementation manners of the present invention or make equivalent replacements. Any modification or equivalent replacement without departing from the spirit and scope of the present invention shall be covered by the protection scope of the claims of the present invention.

Claims

1. A method for irradiance prediction for distributed photovoltaic, characterized in that, The method includes: Clustering distributed photovoltaic sites in the topological map of regional irradiance numerical prediction grid points, and obtaining a clustering square graph corresponding to each clustering cluster; Based on the vertex irradiance numerical prediction values of the grids to which the vertices of the clustering square graph belong in the topological map of regional irradiance numerical prediction grid points, determining the irradiance numerical prediction values of the vertices of the clustering square graph; Based on the irradiance numerical prediction values of the vertices of the clustering square graph, determining the irradiance numerical prediction values of the distributed photovoltaic sites in each clustering cluster; The calculation formula for the irradiance numerical prediction values of the distributed photovoltaic sites in each clustering cluster is as follows: In the above formula, Z is the numerical prediction value of the irradiance of the distributed photovoltaic sites in the clustering cluster, and Z j is the numerical prediction value of the irradiance at vertex j of the clustering square diagram corresponding to the clustering cluster, and Q j is the weight of vertex j of the clustering square diagram corresponding to the clustering cluster; The process of obtaining the weight of vertex j of the clustering square graph corresponding to the clustering cluster includes: Solving the pre-constructed maximum correlation model to obtain the weight of vertex j of the clustering square graph corresponding to the clustering cluster; The pre-constructed maximum correlation model includes: an objective function with the maximum correlation coefficient between the irradiance numerical prediction value of the distributed photovoltaic site in the clustering cluster and the measured total power and the meteorological monitoring irradiance respectively as the target, and constraint conditions configured for the irradiance prediction of distributed photovoltaics; The calculation formula for the objective function is as follows: In the above formula, r ZP is the correlation coefficient between the numerical prediction value of the irradiance of the distributed PV sites in the clustering cluster and the measured total power, and r ZR is the correlation coefficient between the numerical prediction value of the irradiance of the distributed PV sites in the clustering cluster and the meteorological monitoring irradiance; The calculation formula for the correlation coefficient between the irradiance numerical prediction value of the distributed photovoltaic site in the clustering cluster and the measured total power is as follows: The calculation formula for the correlation coefficient between the irradiance numerical prediction value of the distributed photovoltaic site in the clustering cluster and the meteorological monitoring irradiance is as follows: In the above formula, Z is the irradiance numerical prediction value of the distributed photovoltaic site in the clustering cluster, P is the measured total power, R is the meteorological monitoring irradiance, E is the mathematical expectation function, D is the variance function, μ1 is the mean value of the irradiance numerical prediction values of the distributed photovoltaic sites in the clustering cluster, μ2 is the mean value of the measured total power, and μ3 is the mean value of the meteorological monitoring irradiance.

2. The method according to claim 1, characterized in that, The clustering of distributed photovoltaic sites in the topological map of regional irradiance numerical prediction grid points includes: Taking the meteorological monitoring device as the clustering center in the topological map of regional irradiance numerical prediction grid points, and clustering the distributed photovoltaic sites based on the distance between the clustering center and the distributed photovoltaic sites.

3. The method according to claim 2, wherein There is no intersection between the clustering square graphs corresponding to each clustering cluster. If there are distributed photovoltaic sites not covered by the clustering square graph, then the distributed photovoltaic sites not covered by the clustering square graph are classified into the clustering square graph with the closest distance to it.

4. The method according to claim 1, wherein The determining of the irradiance numerical prediction values of the vertices of the clustering square graph based on the vertex irradiance numerical prediction values of the grids to which the vertices of the clustering square graph belong in the topological map of regional irradiance numerical prediction grid points includes: When the vertex of the clustering square graph is located at the vertex of the grid in the topological map of the regional irradiance numerical prediction grid point to which it belongs, the irradiance numerical prediction value of the vertex of the clustering square graph is the irradiance numerical prediction value of the vertex of the grid in the topological map of the regional irradiance numerical prediction grid point to which it belongs; When the vertex of the clustering square graph is located on the boundary of the grid in the topological map of the regional irradiance numerical prediction grid point to which it belongs, based on the irradiance numerical prediction values of the two endpoints of the boundary, using a linear interpolation algorithm to determine the irradiance numerical prediction value of the vertex of the clustering square graph; When the vertex of the clustering square graph is within the grid in the irradiance numerical prediction grid point topology graph of its affiliated area, based on the irradiance numerical prediction values of the four vertices of the grid, the inverse distance squared interpolation algorithm is used to determine the irradiance numerical prediction value of the vertex of the clustering square graph.

5. The method according to claim 4, wherein The calculation formula for determining the irradiance numerical prediction value of the vertex of the clustering square graph by using the linear interpolation algorithm is as follows: In the above formula, Z B is the predicted irradiance value of vertex B of the clustering square diagram, Za is the predicted irradiance value of endpoint a of the boundary, Z b is the predicted irradiance value of endpoint b of the boundary, d ba is the distance between vertex B of the clustering square diagram and endpoint a of the boundary, d bb is the distance between vertex B of the clustering square diagram and endpoint b of the boundary.

6. The method according to claim 4, wherein The calculation formula for determining the irradiance numerical prediction value of the vertex of the clustering square graph by using the inverse distance squared interpolation algorithm is as follows: In the above formula, Z A is the predicted irradiance value of vertex A of the clustering square diagram, and E i is the weight of vertex i of the grid in the topological diagram of the irradiance numerical prediction grid points in the area to which the vertex of the clustering square diagram belongs, and Z i is the predicted irradiance value of vertex i of the grid in the topological diagram of the irradiance numerical prediction grid points in the area to which the vertex of the clustering square diagram belongs.

7. The method according to claim 6, wherein The calculation formula for the weight of vertex i of the grid in the irradiance numerical prediction grid point topology graph of the area to which the vertex of the clustering square graph belongs is as follows: In the above formula, d ai is the distance between vertex A of the clustering square diagram and vertex i of the grid in the irradiance numerical prediction grid diagram of the region to which the vertex of the clustering square diagram belongs.

8. The method according to claim 1, wherein The mathematical model of the constraint condition is: In the above formula, Q A , Q B , Q C , Q D are the weights of the four vertices of the clustering square diagram corresponding to the clustering cluster, respectively.

9. An irradiance prediction device for distributed photovoltaic, characterized in that, The device includes: A clustering module, configured to cluster distributed photovoltaic sites in the irradiance numerical prediction grid point topology graph of the area and obtain a clustering square graph corresponding to each clustering cluster; A first determination module, configured to determine the irradiance numerical prediction value of each vertex of the clustering square graph based on the irradiance numerical prediction values of the vertices of the grid to which each vertex of the clustering square graph belongs in the irradiance numerical prediction grid point topology graph of the area; A second determination module, configured to determine the irradiance numerical prediction value of the distributed photovoltaic sites in each clustering cluster based on the irradiance numerical prediction values of each vertex of the clustering square graph; The calculation formula for the irradiance numerical prediction value of the distributed photovoltaic sites in each clustering cluster is as follows: In the above formula, Z is the numerical prediction value of the irradiance of the distributed PV sites in the clustering cluster, and Z j is the numerical prediction value of the irradiance at vertex j of the clustering square diagram corresponding to the clustering cluster, and Q j is the weight of vertex j of the clustering square diagram corresponding to the clustering cluster; The process of obtaining the weight of vertex j of the clustering square graph corresponding to the clustering cluster includes: Solving a pre-constructed maximum correlation model to obtain the weight of vertex j of the clustering square graph corresponding to the clustering cluster; The pre-constructed maximum correlation model includes: an objective function with the maximum correlation coefficient between the irradiance numerical prediction value of the distributed photovoltaic sites in the clustering cluster and the measured total power and the meteorological monitoring irradiance respectively as the objective, and constraint conditions configured for the irradiance prediction of distributed photovoltaics; The calculation formula for the objective function is as follows: In the above formula, r ZP is the correlation coefficient between the numerical prediction value of the irradiance of the distributed PV sites in the clustering cluster and the measured total power, and r ZR is the correlation coefficient between the numerical prediction value of the irradiance of the distributed PV sites in the clustering cluster and the meteorological monitored irradiance; The calculation formula for the correlation coefficient between the irradiance numerical prediction value of the distributed photovoltaic sites in the clustering cluster and the measured total power is as follows: The calculation formula for the correlation coefficient between the irradiance numerical prediction value of the distributed photovoltaic sites in the clustering cluster and the meteorological monitoring irradiance is as follows: In the above formula, Z is the irradiance numerical prediction value of the distributed photovoltaic sites in the clustering cluster, P is the measured total power, R is the meteorological monitoring irradiance, E is the mathematical expectation function, D is the variance function, μ1 is the mean value of the irradiance numerical prediction values of the distributed photovoltaic sites in the clustering cluster, μ2 is the mean value of the measured total power, and μ3 is the mean value of the meteorological monitoring irradiance.

10. The device according to claim 9, characterized in that, The clustering module is specifically configured to: Take the meteorological monitoring device as the clustering center in the irradiance numerical prediction grid point topology graph of the area, and cluster the distributed photovoltaic sites based on the distance between the clustering center and the distributed photovoltaic sites.

11. The device according to claim 10, characterized in that There is no intersection between the clustering square graphs corresponding to each clustering cluster. If there are distributed photovoltaic sites not covered by the clustering square graph, then the distributed photovoltaic sites not covered by the clustering square graph are classified into the clustering square graph with the closest distance to it.

12. The device according to claim 9, characterized in that, The first determination module is specifically configured to: When the vertex of the clustering square graph is located at the vertex of the grid in the irradiance numerical prediction grid point topology map of its affiliated area, the irradiance numerical prediction value of the vertex of the clustering square graph is the irradiance numerical prediction value of the vertex of the grid in the irradiance numerical prediction grid point topology map of its affiliated area; When the vertex of the clustering square graph is located on the boundary of the grid in the irradiance numerical prediction grid point topology map of its affiliated area, based on the irradiance numerical prediction values of the two endpoints of the boundary, a linear interpolation algorithm is used to determine the irradiance numerical prediction value of the vertex of the clustering square graph; When the vertex of the clustering square graph is located inside the grid in the irradiance numerical prediction grid point topology map of its affiliated area, based on the irradiance numerical prediction values of the four vertices of the grid, an inverse distance squared interpolation algorithm is used to determine the irradiance numerical prediction value of the vertex of the clustering square graph.

13. The device according to claim 12, wherein The calculation formula for determining the irradiance numerical prediction value of the vertex of the clustering square graph using the linear interpolation algorithm is as follows: In the above formula, Z B is the predicted irradiance value of vertex B of the clustering square diagram, Za is the predicted irradiance value of endpoint a of the boundary, Z b is the predicted irradiance value of endpoint b of the boundary, d ba is the distance between vertex B of the clustering square diagram and endpoint a of the boundary, d bb is the distance between vertex B of the clustering square diagram and endpoint b of the boundary.

14. The device according to claim 12, characterized in that, The calculation formula for determining the irradiance numerical prediction value of the vertex of the clustering square graph using the inverse distance squared interpolation algorithm is as follows: In the above formula, Z A is the predicted irradiance value of vertex A of the clustering square diagram, and E i is the weight of vertex i of the grid in the topological diagram of the irradiance numerical prediction grid points in the area to which the vertex of the clustering square diagram belongs, and Z i is the predicted irradiance value of vertex i of the grid in the topological diagram of the irradiance numerical prediction grid points in the area to which the vertex of the clustering square diagram belongs.

15. The device according to claim 14, wherein, The calculation formula for the weight of vertex i of the grid in the irradiance numerical prediction grid point topology map of the area to which the vertex of the clustering square graph belongs is as follows: In the above formula, d ai is the distance between vertex A of the clustering square diagram and vertex i of the grid in the irradiance numerical prediction grid diagram of the area to which the vertex of the clustering square diagram belongs.

16. The device according to claim 9, characterized in that, The mathematical model of the constraint condition is: In the above formula, Q A , Q B , Q C , Q D are the weights of the four vertices of the clustering square diagram corresponding to the clustering cluster, respectively.

17. A computer device, characterized in that, Including: One or more processors; The processor is used to store one or more programs; When the one or more programs are executed by the one or more processors, the irradiance prediction method for distributed photovoltaic as described in any one of claims 1 to 8 is implemented.

18. A computer-readable storage medium, characterized in that, There is a computer program stored thereon, and when the computer program is executed, the irradiance prediction method for distributed photovoltaic as described in any one of claims 1 to 8 is implemented.

Images