A power generation prediction system and method for photovoltaic power stations in mountainous areas

Abstract

This application discloses a power generation prediction system and method for a photovoltaic power station in mountainous areas, relating to the field of photovoltaic power generation technology. It solves the technical problem that existing technologies, which predict power generation using meteorological forecast data and artificial intelligence models, are not applicable to photovoltaic power stations located in complex terrain environments. This application establishes environmental time-series data for each photovoltaic panel based on irradiance time-series data and meteorological time-series data. This environmental time-series data is input into a power prediction model to obtain the power generation of the target power station. The irradiance time-series data in this application is used to determine whether each photovoltaic panel in the target power station is shaded, while the meteorological time-series data is the meteorological data for each photovoltaic panel at a given time. Combining these two data allows for the determination of the environmental time-series data for each photovoltaic panel. Based on the environmental time-series data, the power generation environment of the photovoltaic panels at each time can be determined, thereby determining the power generation of the target power station at that time. This application fully considers the influence of micro-meteorological effects to improve the prediction accuracy of power generation.

Description

Technical Field

[0001] This application belongs to the field of photovoltaic power generation and relates to the power generation prediction technology of photovoltaic power stations, specifically a power generation prediction system and method for a photovoltaic power station used in mountainous areas. Background Technology

[0002] With the continuous development of clean energy, the proportion of photovoltaic power generation is growing rapidly. Since photovoltaic power plants require a large amount of land, hillsides with abundant sunshine that are difficult to utilize are gradually being used to build photovoltaic power plants.

[0003] The mountainous terrain is highly undulating, and the photovoltaic arrays are scattered and complexly zoned. The installation angles of photovoltaic power stations vary in different areas, making them easily shaded by changes in the sun's azimuth, with varying shaded areas. Furthermore, micro-meteorological effects are highly likely to occur on mountain slopes, causing local weather conditions to differ significantly from the surrounding areas, thus impacting the surrounding regions. Existing methods, which predict the power generation of photovoltaic power stations based on forecasted meteorological data, are clearly unsuitable for predicting the power generation of photovoltaic power stations in mountainous terrain.

[0004] This application provides a power generation prediction system and method for a photovoltaic power station in mountainous areas to solve the above-mentioned technical problems. Summary of the Invention

[0005] This application aims to solve at least one of the technical problems existing in the prior art; to this end, this application proposes a power generation prediction system and method for mountain photovoltaic power stations, which solves the technical problem that the existing technology of predicting power generation through meteorological forecast data and artificial intelligence models cannot be applied to photovoltaic power stations set up in complex terrain environments.

[0006] To achieve the above objectives, the first aspect of this application provides a method for predicting the power generation of a photovoltaic power station in mountainous areas, comprising:

[0007] The location feature information of the target power station is collected and acquired. Based on the location feature information, the target power station is modeled, and the solar motion trajectory is simulated to determine the time series data of the irradiance of each photovoltaic panel in the target power station. Among them, the location feature information includes the location, altitude, angle of each photovoltaic panel and the surrounding environment data.

[0008] The area covered by the micro-meteorological phenomena in the target power station area is identified and marked as meteorological region one; meteorological forecast data of the target power station and micro-meteorological data of meteorological region one are obtained; meteorological time series data of each photovoltaic panel are constructed based on the meteorological forecast data and micro-meteorological data.

[0009] Environmental time-series data for each photovoltaic panel is established based on irradiance time-series data and meteorological time-series data. The environmental time-series data is then input into the power prediction model to obtain the power generation capacity of the target power plant. The power prediction model is constructed based on an artificial intelligence model.

[0010] Preferably, modeling the target power plant based on location feature information includes:

[0011] After preprocessing the location feature information, it is input into the modeling software to construct a virtual model of the target power station; the modeling software includes Pvsyst or SketchUp.

[0012] By simulating the solar trajectory in different seasons, the impact on each photovoltaic panel in the virtual model during the solar motion is identified, and the impact is converted into time-series data of the irradiance of each photovoltaic panel.

[0013] Preferably, the coverage area for detecting micrometeorological phenomena in the area where the target power station is located includes:

[0014] Historical meteorological data of the target power station is extracted; the historical meteorological data is classified according to meteorological elements to obtain data of each element; among them, meteorological elements include temperature, wind speed, humidity and solar irradiance;

[0015] If a micro-meteorological phenomenon exists, its coverage area is determined based on the changing trends of the data of each element; if yes, the coverage area of ​​the micro-meteorological phenomenon is determined based on the changing trends; otherwise, it is not necessary to determine the coverage area.

[0016] Preferably, determining the existence of micrometeorological phenomena based on the changing trends of various element data includes:

[0017] Three-dimensional change maps of corresponding meteorological elements are built based on element data, and abnormal areas in the three-dimensional change maps are identified respectively; the three-dimensional change maps are used to represent the changes of meteorological elements at different locations.

[0018] When at least one meteorological element corresponds to an overlapping anomalous area, it is determined that there is a micrometeorological phenomenon in the area where the target power station is located.

[0019] Preferably, the coverage area of ​​micrometeorological phenomena is determined based on the changing trend, including:

[0020] At least one target element is selected from meteorological elements based on its impact weight on photovoltaic power generation; the target elements are selected in descending order of their impact weight.

[0021] The abnormal area of ​​the target element is taken as the coverage area of ​​the micro-meteorological phenomenon, or the combined abnormal areas corresponding to several target elements are taken as the coverage area of ​​the micro-meteorological phenomenon.

[0022] Preferably, meteorological time-series data for each photovoltaic panel is constructed based on meteorological forecast data and micrometeorological data, including:

[0023] Extract meteorological forecast data and micrometeorological data;

[0024] The locations of the photovoltaic panels in the target site are taken as the target locations, and the meteorological data corresponding to the target locations are extracted from meteorological forecast data or micro-meteorological data.

[0025] Meteorological time-series data is constructed based on meteorological data and their corresponding times, and then linked to photovoltaic panels.

[0026] Preferably, environmental time-series data for each photovoltaic panel is established based on irradiance time-series data and meteorological time-series data, including:

[0027] Extract the irradiance time-series data and environmental time-series data of each photovoltaic panel in the target power plant; where the irradiance time-series data refers to the amount of sunlight that the photovoltaic panel can receive at each time.

[0028] By aligning the irradiance time-series data and the environmental time-series data, the environmental time-series data of the photovoltaic panel is obtained.

[0029] Preferably, before obtaining the environmental time-series data of the photovoltaic panels, the presence of clouds or fog that could affect the target power plant is detected;

[0030] If it exists, predict the cloud and fog movement trajectory, and adjust the time-aligned data based on the cloud and fog movement trajectory to obtain the environmental time-series data of each photovoltaic panel;

[0031] If it does not exist, no adjustments are made, and the environmental time series data of each photovoltaic panel is obtained directly.

[0032] Preferably, inputting environmental time-series data into the power prediction model includes:

[0033] Several prediction times are determined; wherein, the prediction times are extracted from the set prediction time period at set intervals;

[0034] Data on photovoltaic panels at several prediction times are extracted from environmental time-series data and labeled as basic model data; the basic model data of each photovoltaic panel are integrated into model input data.

[0035] Input the model input data into the power prediction model to obtain the power generation of the target power plant at several prediction times; construct the power prediction curve based on the power generation at several prediction times.

[0036] The second aspect of this application provides a power generation prediction system for a photovoltaic power station in mountainous areas, including a power prediction module and a data acquisition module connected thereto;

[0037] Data acquisition module: used to collect and acquire the location feature information of the target power station; the location feature information includes the location, altitude, angle of each photovoltaic panel and surrounding environmental data;

[0038] Power prediction module: used to model the target power plant based on location feature information and simulate the solar trajectory to determine the time series data of irradiance of each photovoltaic panel in the target power plant;

[0039] The area covered by micro-meteorological phenomena in the target power plant area is identified and marked as meteorological region one; meteorological forecast data for the target power plant and micro-meteorological data for meteorological region one are acquired; based on the meteorological forecast data and micro-meteorological data, meteorological time-series data for each photovoltaic panel are constructed; and,

[0040] Environmental time-series data for each photovoltaic panel is established based on irradiance time-series data and meteorological time-series data. The environmental time-series data is then input into the power prediction model to obtain the power generation capacity of the target power plant. The power prediction model is constructed based on an artificial intelligence model.

[0041] Compared with the prior art, the beneficial effects of this application are:

[0042] 1. This application models a target power station based on location feature information, simulates the solar trajectory to determine the irradiance time-series data of each photovoltaic panel in the target power station; predicts and obtains meteorological forecast data and micro-meteorological data of meteorological region 1 for the target power station; constructs meteorological time-series data for each photovoltaic panel based on the meteorological forecast data and micro-meteorological data; establishes environmental time-series data for each photovoltaic panel based on the irradiance time-series data and meteorological time-series data, and inputs the environmental time-series data into the power prediction model to obtain the power generation of the target power station; the irradiance time-series data in this application is used to determine whether each photovoltaic panel in the target power station is shaded, and the meteorological time-series data is the meteorological data of each photovoltaic panel at a given time. The combination of the two can determine the environmental time-series data of each photovoltaic panel, and the power generation environment of the photovoltaic panel at each time can be determined based on the environmental time-series data, thereby determining the power generation of the target power station at that time. This application fully considers the influence of micro-meteorological effects to improve the prediction accuracy of power generation.

[0043] 2. Before obtaining the environmental time-series data of the photovoltaic panels, this application detects whether there are clouds or fog that may affect the target power station. If they exist, the application predicts the movement trajectory of the clouds and fog and adjusts the data obtained in time alignment based on the movement trajectory of the clouds and fog to obtain the environmental time-series data of each photovoltaic panel. If they do not exist, no adjustment is made, and the environmental time-series data of each photovoltaic panel is obtained directly. This application considers the randomness of clouds and fog and detects whether there are clouds or fog that may affect the target power station before determining the environmental time-series data. If they exist, the application adjusts the data based on the predicted movement trajectory of the clouds and fog. This can solve the impact of randomly occurring clouds and fog in the mountains on the target power station, thereby improving the prediction accuracy of the power generation of the target power station. Attached Figure Description

[0044] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0045] Figure 1 This is a schematic diagram of the method steps for predicting power generation in Embodiment 1 of this application;

[0046] Figure 2 This is a schematic diagram of the power generation prediction system in Embodiment 1 of this application;

[0047] Figure 3 This is a schematic diagram of the method steps for adjusting the motion trajectory of clouds and fog in Embodiment 2 of this application. Detailed Implementation

[0048] The technical solutions of this application will be clearly and completely described below with reference to the embodiments. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0049] Setting up a photovoltaic power station on a hillside is subject to additional factors affecting its power generation compared to setting it on a plain: 1) Topography and orientation: Different hillside orientations, slopes, and the safe angle of photovoltaic modules will affect the amount of sunlight received; 2) Shading area and size: Shading from the mountain and between photovoltaic modules will lead to a loss of power generation; 3) Local climate and micrometeorological effects: Local climate and micrometeorological effects are more significant in mountainous areas, such as valley winds and airflow changes caused by topography. These factors will affect parameters such as solar irradiance and temperature, and thus affect power generation.

[0050] Most existing technical solutions predict the power generation of photovoltaic power plants by forecasting future weather data and combining the sun's trajectory with the installation angle of the photovoltaic panels. However, these existing solutions do not disclose solutions for the factors mentioned above. To accurately predict the power generation of photovoltaic power plants used in mountainous areas, this application provides a power generation prediction system and method for such plants.

[0051] Example 1:

[0052] Please see Figures 1-2 The first aspect of this application provides a method for predicting the power generation of a photovoltaic power station in mountainous areas, including:

[0053] The system collects and acquires the location feature information of the target power station, models the target power station based on the location feature information, and simulates the solar motion trajectory to determine the irradiance time series data of each photovoltaic panel in the target power station; detects the coverage area of ​​micro-meteorological phenomena in the area where the target power station is located and marks it as meteorological region one; predicts and acquires meteorological forecast data of the target power station and micro-meteorological data of meteorological region one; constructs meteorological time series data of each photovoltaic panel based on meteorological forecast data and micro-meteorological data; establishes environmental time series data of each photovoltaic panel based on irradiance time series data and meteorological time series data, and inputs the environmental time series data into the power prediction model to obtain the power generation of the target power station.

[0054] To address the issue of shading of photovoltaic panels in a solar power plant caused by the complex terrain of mountainous areas, this embodiment first collects location feature information of the target power plant. This location feature information mainly includes the location, altitude, and angle of each photovoltaic panel, as well as surrounding environmental data. The surrounding environmental data mainly refers to factors that affect the photovoltaic panels' ability to receive sunlight, such as tall trees, surrounding rocks, and cliffs. Based on the location feature information, a model of the target power plant can be created. Then, the solar trajectory can be simulated to determine which photovoltaic panels will receive sunlight at what times, laying a data foundation for subsequent prediction of the target power plant's power generation capacity.

[0055] To address the impact of micro-meteorological effects on the target power plant, this embodiment determines whether the area where the target power plant is located is affected by micro-meteorological effects. If affected, the coverage area of ​​the micro-meteorological effects is identified, and then meteorological data of the coverage area is predicted to assist in predicting the power generation of the target power plant. By considering the impact of micro-meteorological effects, the prediction accuracy of power generation can be improved.

[0056] Next, the specific implementation process of this embodiment will be described in detail.

[0057] First, a model is created for the target power plant based on its location features, including:

[0058] After preprocessing the location feature information, it is input into the modeling software to construct a virtual model of the target power plant. By simulating the sun's trajectory in different seasons, the impact of the sun's movement on each photovoltaic panel in the virtual model is identified, and the impact is converted into time-series irradiance data for each photovoltaic panel. Many relevant modeling software programs are available, such as Pvsyst or SketchUp.

[0059] For example, using Pvsyst software to obtain irradiance time-series data:

[0060] 1. Setting up project sites and meteorological data: After creating a new project in the software, select the site file for the target power station; alternatively, you can create a new site directly. Then, enter the location information of the target power station, such as latitude and longitude, and import it after confirmation. Next, import the relevant information of the surrounding environmental data of the target power station determined by the field survey into the software to build a virtual model of the target power station.

[0061] 2. Simulate the sun's trajectory: After entering the simulation interface, set the corresponding parameters according to the season and time range to be simulated. After running the simulation, you can simulate the sun's trajectory in different seasons.

[0062] 3. Analysis of the impact: After the simulation is completed, the software will generate a detailed report document; the impact of each photovoltaic panel can be identified through the report document; the impact of each photovoltaic panel is converted into irradiance time series data, which serves as a typical data basis for the prediction of the power generation of the target power plant in the future.

[0063] It should be noted that the irradiance time-series data mainly includes the time period of day during which the corresponding photovoltaic panels receive sunlight. This time period may be a complete period or multiple discontinuous periods. Moreover, the irradiance time-series data is closely related to the location of the target power plant and its surrounding environment. Since the location and environment generally do not change without human intervention, the irradiance time-series data of each photovoltaic panel in the target power plant is also relatively stable due to the relatively stable location characteristics. The data obtained after one modeling and simulation can be used for a long time. When the data of the target power plant or its surrounding environment changes, targeted modeling and simulation can be performed.

[0064] Secondly, the coverage area of ​​micro-meteorological phenomena in the area where the target power station is located is detected, including: extracting historical meteorological data of the target power station; classifying the historical meteorological data according to meteorological elements to obtain data of each element; judging whether micro-meteorological phenomena exist based on the changing trends of each element data; if yes, then determining the coverage area of ​​micro-meteorological phenomena based on the changing trends; if not, then it is not necessary to determine the coverage area.

[0065] Micrometeorological effects refer to meteorological characteristics on a small spatial scale, ranging from a few meters to a few kilometers. These characteristics may make local meteorological conditions significantly different from the surrounding areas, but they do not cause major changes to large-scale climate characteristics. Micrometeorological effects can affect temperature, humidity, wind speed, and solar radiation, and changes in these parameters can affect the power generation capacity of photovoltaic power plants.

[0066] Micrometeorological effects can be determined by analyzing historical meteorological data of the area where the target power plant is located. This historical meteorological data can be obtained through surveys conducted before the construction of the target power plant. Based on the characteristics of micrometeorological effects, the existence of micrometeorological phenomena can be determined by observing the changing trends of various meteorological elements. After obtaining the historical meteorological data, it is first classified according to meteorological elements, resulting in several data groups, each corresponding to a meteorological element.

[0067] Determining the existence of micro-meteorological phenomena based on the changing trends of various element data includes:

[0068] Based on the element data, a three-dimensional change map of the corresponding meteorological elements is established, and abnormal areas in the three-dimensional change map are identified respectively; when the abnormal areas corresponding to at least one meteorological element overlap, it is determined that there are micrometeorological phenomena in the area where the target power station is located.

[0069] A 3D variation plot is used to represent the changes of meteorological elements at different locations. This 3D variation plot can be generated by importing data sets into MATLAB software. Each 3D variation plot is associated with a meteorological element. In the 3D variation plot, the X and Y axes represent the location, and the Z axis represents the specific value of the meteorological element.

[0070] For example, temperature is used to illustrate the construction of a three-dimensional change graph and the judgment of microclimate phenomena:

[0071] Extract temperature-related data sets from historical meteorological data. These data sets include temperature values ​​and their corresponding location information, which can be latitude and longitude or determined according to other set coordinate systems.

[0072] Import the data set into MATLAB software and call the relevant image generation function to generate a three-dimensional temperature change graph. If there is an area in the three-dimensional change graph that is significantly different from the surrounding area, the abnormal area is identified (MATLAB software has functions to identify abrupt change points). At this time, it can also be determined that there are micrometeorological phenomena in the area where the target power station is located.

[0073] It is worth noting that the identification of anomalous areas is mainly based on whether there are sudden temperature changes. If the temperature changes abruptly at certain points, and the degree of change exceeds a set threshold, then the area enclosed by these points is considered an anomalous area. The threshold can be set based on the experience of relevant meteorological experts.

[0074] When confirming the existence of a micrometeorological phenomenon, it is also necessary to determine the area it covers. In this example, by identifying which meteorological factors have a significant impact on photovoltaic power generation, the coverage area is determined based on the corresponding anomalous areas. Specifically:

[0075] Select at least one target element from meteorological elements based on the weight of its impact on photovoltaic power generation; take the anomalous area of ​​the target element as the coverage area of ​​the micro-meteorological phenomenon, or combine the anomalous areas corresponding to several target elements as the coverage area of ​​the micro-meteorological phenomenon.

[0076] The influence weights of each meteorological element on photovoltaic power generation can be determined using principal component analysis and equal-weighted analysis. The meteorological elements are ranked from largest to smallest influence weight, and the element with the largest influence weight is selected as the target element. If there is only one target element, its anomalous area is taken as the coverage area of ​​the micro-meteorological phenomenon; if there are multiple target elements, the anomalous areas of all target elements are merged as the coverage area of ​​the micro-meteorological phenomenon.

[0077] The reason for determining the target elements using influence weights in this embodiment is that some meteorological elements in micro-meteorological effects do not affect the power generation of photovoltaic power plants, or have a minor impact on power generation, and are therefore not analyzed to avoid increasing the data processing workload. For example, micro-meteorological effects may affect the humidity in a region, but this humidity mainly affects the ground and vegetation, and its impact on power generation is relatively small compared to temperature; wind speed in a region may also vary significantly, but its impact on power generation is also relatively small. Considering the data processing workload, meteorological elements with minor impacts on power generation, such as humidity and wind speed, can be analyzed selectively based on the influence weights of specific regions.

[0078] After determining that a micrometeorological effect exists in the area where the target power station is located, meteorological forecast data and micrometeorological data for the area are predicted based on historical meteorological data. When obtaining meteorological forecast data and micrometeorological data, the photovoltaic panel should be used as a benchmark, that is, the historical meteorological data corresponding to the photovoltaic panel should be obtained. The meteorological forecast data is obtained by using the historical meteorological data and the meteorological forecast model. If the photovoltaic panel is located within the coverage area of ​​the micrometeorological phenomenon, then the meteorological forecast data is the micrometeorological data.

[0079] It should be noted that the meteorological forecast data and micro-meteorological data can be obtained through time series models, such as autoregressive moving average models or long short-term memory network models. These models for forecasting meteorological data have been publicly disclosed in existing schemes, and the specific training and forecasting processes will not be elaborated here.

[0080] Next, it is necessary to construct the meteorological time-series data for each photovoltaic panel in the target power plant, specifically:

[0081] Extract meteorological forecast data and micro-meteorological data; sequentially take the location of the photovoltaic panels in the target site as the target location, and extract the meteorological data corresponding to the target location from the meteorological forecast data or micro-meteorological data; construct meteorological time series data based on the meteorological data and its corresponding time, and associate it with the photovoltaic panels.

[0082] Extract the meteorological forecast data and micro-meteorological data obtained from the aforementioned scheme, and then use the location of each photovoltaic panel as the target location; extract the meteorological data of the target location from the meteorological forecast data and micro-meteorological data, and then integrate them in chronological order to obtain meteorological time series data.

[0083] For example, suppose the target power plant has photovoltaic panel A and photovoltaic panel B, photovoltaic panel A is in the normal area, and photovoltaic panel B is in the area covered by micro-meteorological phenomena;

[0084] Taking the location of photovoltaic panel A as the target location, meteorological time series data corresponding to the target location is extracted from meteorological forecast data, that is, meteorological data corresponding to each time moment; taking the location of photovoltaic panel B as the target location, meteorological time series data corresponding to the target location is extracted from micro-meteorological data.

[0085] When setting up a photovoltaic power station on a hillside, the photovoltaic panels may be located in different areas and far apart. Therefore, some panels may not be within the coverage area of ​​micro-meteorological phenomena, while others may be. Furthermore, the location of each panel is different, resulting in different meteorological time-series data for each panel. Changes in various meteorological factors within the meteorological time-series data affect the power generation capacity of the photovoltaic panels, thereby influencing the overall power generation capacity of the photovoltaic power station.

[0086] It is necessary to combine meteorological time-series data with irradiance time-series data to determine the environmental time-series data of each photovoltaic panel. Specifically, the irradiance time-series data and environmental time-series data of each photovoltaic panel in the target power station are extracted; the irradiance time-series data and environmental time-series data are time-aligned to obtain the environmental time-series data of the photovoltaic panels.

[0087] Irradiance time-series data refers to the amount of sunlight a photovoltaic (PV) panel can receive at various times each day. During the daily solar cycle, PV panels may not receive sunlight at certain times due to shading from surrounding environmental data; therefore, they cannot generate electricity. Meteorological time-series data does not consider the issue of shading when acquiring PV panels; it indicates whether the PV panels can receive sunlight at a given moment without considering the influence of surrounding environmental data.

[0088] By combining irradiance time-series data and meteorological time-series data, the system sequentially determines whether the photovoltaic panels are shaded at each time point using the irradiance time-series data. If the panels are not shaded at a given time, the corresponding meteorological data for that time point is extracted from the meteorological time-series data. Therefore, after processing, the meteorological data for each time point, i.e., environmental time-series data, can be obtained. It should be noted that if the photovoltaic panels are shaded at a certain time, meteorological data does not need to be extracted; the corresponding meteorological data can be set to empty, and this time point will not be considered in subsequent power generation prediction.

[0089] It is worth noting that the photovoltaic panels may be partially shaded. If the remaining parts of the panels can still generate electricity, then it is necessary to match them with meteorological data at various times. However, when subsequently predicting the power generation capacity of the target power plant, it is necessary to incorporate the actual power generation area of ​​each photovoltaic panel in order to accurately predict the power generation capacity of the target power plant.

[0090] By inputting the environmental time-series data of each photovoltaic panel in the target power plant into the power prediction model, the power generation of the target power plant at each time point can be obtained. Before inputting, the environmental time-series data needs to be processed. After normalizing the environmental time-series data, the label (which can be a number) of the photovoltaic panel is inserted. If part of the photovoltaic panel is shaded, the actual working area of ​​the photovoltaic panel also needs to be inserted.

[0091] The power prediction model is trained based on an artificial intelligence model, including:

[0092] Construct an artificial intelligence model; train the artificial intelligence model using standard training data, and label the trained artificial intelligence model as a power prediction model; wherein, the artificial intelligence model includes a BP neural network model or an RBF neural network model;

[0093] Standard training data can be obtained through actual measurements or simulations. The standard training data mainly includes the power generation capacity of the target power plant, as well as the meteorological data and actual working area corresponding to each photovoltaic panel. After normalizing the meteorological data and actual working area of ​​each photovoltaic panel, a label for each photovoltaic panel is inserted. Thus, at any given time, each photovoltaic panel corresponds to a data record containing meteorological data, actual working area, and label. The data corresponding to each photovoltaic panel at the same time are integrated into the model input data, and the power generation capacity of the target power plant at the corresponding time is integrated into the model output data. The constructed artificial intelligence model is trained using the model input and model output data to obtain the power prediction model. The model training process will not be elaborated here.

[0094] Environmental time-series data is input into the power prediction model, including:

[0095] Determine several prediction time points; extract data of photovoltaic panels at several prediction time points from environmental time series data and mark them as basic model data; integrate the basic model data of each photovoltaic panel into model input data; input the model input data into the power prediction model to obtain the power generation of the target power plant at several prediction time points; construct a power prediction curve based on the power generation at several prediction time points.

[0096] The predicted time is extracted from the set prediction period at set intervals. For example, if the prediction period is the next week and the set interval is one hour, then several predicted times can be extracted from the prediction period according to the set interval.

[0097] After obtaining the power generation at several predicted times, a power prediction curve can be fitted and constructed. This power prediction curve represents the change in power generation during the predicted period and can be used to assist staff in dispatching the distribution network. Of course, it can also be used to perform integral calculations to determine the power generation of the target power plant in a certain period.

[0098] The second aspect of this application provides a power generation prediction system for a photovoltaic power station in mountainous areas, including a power prediction module and a data acquisition module connected thereto;

[0099] Data acquisition module: used to collect and acquire the location feature information of the target power station; the location feature information includes the location, altitude, angle of each photovoltaic panel and surrounding environmental data;

[0100] Power prediction module: used to model the target power plant based on location feature information and simulate the solar trajectory to determine the time series data of irradiance of each photovoltaic panel in the target power plant;

[0101] The area covered by micro-meteorological phenomena in the target power plant area is identified and marked as meteorological region one; meteorological forecast data for the target power plant and micro-meteorological data for meteorological region one are acquired; based on the meteorological forecast data and micro-meteorological data, meteorological time-series data for each photovoltaic panel are constructed; and,

[0102] Environmental time-series data for each photovoltaic panel is established based on irradiance time-series data and meteorological time-series data. The environmental time-series data is then input into the power prediction model to obtain the power generation capacity of the target power plant. The power prediction model is constructed based on an artificial intelligence model.

[0103] The data acquisition module is connected to the meteorological platform and various types of meteorological sensors.

[0104] Example 2: Compared to Example 1, this example provides a solution to address the impact of clouds on power generation. Please refer to [link / reference]. Figure 3 Before obtaining the environmental time-series data of the photovoltaic panels, it is necessary to detect whether there are clouds or fog that may affect the target power station. If they exist, the movement trajectory of the clouds or fog is predicted, and the data obtained by time alignment is adjusted based on the movement trajectory of the clouds or fog to obtain the environmental time-series data of each photovoltaic panel. If they do not exist, no adjustment is made, and the environmental time-series data of each photovoltaic panel is obtained directly.

[0105] The occurrence of clouds and fog is highly random, making prediction based on historical data difficult. This embodiment first identifies clouds and fog using cloud and fog detection equipment, such as satellite images or other image data placed around the target power plant. Then, it predicts the future movement trajectory of the clouds and fog using an existing cloud and fog trajectory prediction model, determining whether the movement trajectory will affect the photovoltaic panels in the target power plant. If it will have an impact, the environmental time-series data for each photovoltaic panel is adjusted based on the cloud and fog movement trajectory.

[0106] The presence of clouds and fog can block sunlight from reaching solar panels, preventing them from generating electricity. Therefore, clouds and fog can be understood as dynamic obstacles that obstruct solar panels as they move. By predicting and obtaining the movement trajectory of clouds and fog, we can determine when and which solar panel will be affected. The meteorological data for the corresponding moment for each solar panel can be set to empty, indicating that the panel is not generating electricity at that moment.

[0107] This embodiment first detects whether clouds or fog have formed, then analyzes meteorological data to determine if the clouds or fog will affect the target power plant. Next, it uses an existing cloud and fog tracking platform to predict the movement trajectory of the clouds and fog. It should be noted that if clouds or fog appear during the prediction of the target power plant's power generation capacity, the movement trajectory of the clouds and fog can be analyzed in real time to adjust the environmental time-series data for each photovoltaic panel. Alternatively, clouds and fog can be predicted first and then adjusted to correct the prediction results. Cloud and fog tracking platforms can include NVIDIA Earth-2, RayCloudSim, etc.

[0108] It should be noted that if the occurrence of clouds and fog can be accurately predicted, the clouds and fog and their movement trajectory can be predicted first, thereby adjusting the environmental time series data.

[0109] Example 3: Compared to Example 1, this example uses an artificial intelligence model to predict micro-meteorological data.

[0110] First, meteorological forecast data for the area where the target power station is located is obtained through a meteorological platform. Then, the meteorological forecast data is preprocessed and input into a trained artificial intelligence model to obtain meteorological forecast data for the area covered by micro-meteorological phenomena, i.e., micro-meteorological data.

[0111] In this embodiment, the artificial intelligence model is trained using historical meteorological data from the target power station. Meteorological data of the micro-meteorological phenomena coverage area is extracted from the historical meteorological data and used as historical micro-meteorological data. The historical meteorological data is used as the model input data, and the historical micro-meteorological data is used as the model output data to train the artificial intelligence model. The trained artificial intelligence model can then construct a mapping relationship between the meteorological data and the micro-meteorological data.

[0112] Of course, to achieve accurate prediction of micro-meteorological data, several feature points can be selected from the coverage area, and the meteorological data of these feature points can be integrated into micro-meteorological data. The meteorological platform can be a third-party meteorological forecasting platform, or data provided by meteorological stations set up in the area where the target power station is located.

[0113] The above embodiments are only used to illustrate the technical methods of this application and are not intended to limit it. Although this application has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical methods of this application without departing from the spirit and scope of the technical methods of this application.

Claims

1. A method of predicting power generation of a mountain photovoltaic power plant, characterized by, The method comprises the following steps: Collecting position characteristic information of a target power station, modeling the target power station based on the position characteristic information, and simulating a sun motion trajectory to determine irradiance time series data of each photovoltaic panel in the target power station; wherein the position characteristic information comprises the position, altitude, angle and surrounding environment data of each photovoltaic panel; Detecting a coverage area of a microclimate phenomenon in the area where the target power station is located, and marking it as a meteorological area 1; predicting and obtaining meteorological prediction data of the target power station and microclimate data of the meteorological area 1; and constructing meteorological time series data of each photovoltaic panel based on the meteorological prediction data and the microclimate data; The detection of the coverage area of the microclimate phenomenon in the area where the target power station is located comprises: Extracting historical meteorological data of the target power station; classifying the historical meteorological data according to meteorological elements to obtain element data; wherein the meteorological elements include temperature, wind speed, humidity and solar irradiance; Judging whether there is a microclimate phenomenon based on the change trend of each element data; if yes, determining the coverage area of the microclimate phenomenon according to the change trend; if no, no coverage area needs to be determined; The judgment of whether there is a microclimate phenomenon based on the change trend of each element data comprises: Establishing a three-dimensional change graph of the corresponding meteorological element based on the element data, and identifying the abnormal area in the three-dimensional change graph respectively; wherein the three-dimensional change graph is used to represent the change of the meteorological element at different positions; When the abnormal areas corresponding to at least one meteorological element overlap, it is judged that there is a microclimate phenomenon in the area where the target power station is located; The determination of the coverage area of the microclimate phenomenon according to the change trend comprises: Selecting at least one target element from the meteorological elements according to the influence weight on photovoltaic power generation; wherein the target elements are selected in descending order of influence weight; Taking the abnormal area of the target element as the coverage area of the microclimate phenomenon, or taking the combination of the abnormal areas corresponding to several target elements as the coverage area of the microclimate phenomenon; Based on the irradiance time series data and the meteorological time series data, the environmental time series data of each photovoltaic panel is established, the environmental time series data is input into a power prediction model, and the power generation of the target power station is obtained; wherein the power prediction model is constructed based on an artificial intelligence model.

2. The method of claim 1, wherein the method is characterized by: Modeling the target power station based on the position characteristic information comprises: After preprocessing the position characteristic information, input it into modeling software to construct a virtual model of the target power station; wherein the modeling software includes Pvsyst or SketchUp; By simulating the sun motion trajectory in different seasons, the affected situation of each photovoltaic panel in the virtual model during the sun motion is identified, and the affected situation is converted into irradiance time series data of each photovoltaic panel.

3. The method of claim 1, wherein the method is characterized by: Based on the meteorological prediction data and the microclimate data, the meteorological time series data of each photovoltaic panel is constructed, which comprises: Extracting the meteorological prediction data and the microclimate data; Taking the position of the photovoltaic panel in the target power station as a target position in turn, extracting the meteorological data corresponding to the target position from the meteorological prediction data or the microclimate data; Construct weather time series data based on the weather data and its corresponding time, and associate it with the photovoltaic panel.

4. The method of claim 3, wherein the method is characterized by: Establish the environmental time series data of each photovoltaic panel based on the irradiance time series data and the weather time series data, including: Extract the irradiance time series data and the environmental time series data of each photovoltaic panel in the target power station; wherein the irradiance time series data refers to the light that the photovoltaic panel can receive at each time; Time-align the irradiance time series data and the environmental time series data to obtain the environmental time series data of the photovoltaic panel.

5. The method of claim 4, wherein the method further comprises: Before obtaining the environmental time series data of the photovoltaic panel, detect whether there is cloud and fog affecting the target power station; If there is, predict the cloud and fog movement trajectory, and adjust the data obtained in the time alignment based on the cloud and fog movement trajectory to obtain the environmental time series data of each photovoltaic panel; If not, do not adjust, and directly obtain the environmental time series data of each photovoltaic panel.

6. The method of claim 5, wherein the method further comprises: Input the environmental time series data into the power prediction model, including: Determine a number of prediction times; wherein the prediction times are extracted at a set interval from a set prediction period; Extract the data of the photovoltaic panel at a number of prediction times from the environmental time series data, marked as basic model data; integrate the basic model data of each photovoltaic panel into model input data; Input the model input data into the power prediction model to obtain the power generation of the target power station at a number of prediction times; based on the power generation at a number of prediction times, construct a power prediction curve.

7. A system for predicting the power generation of a mountain photovoltaic power station, for performing the method for predicting the power generation of a mountain photovoltaic power station according to any one of claims 1 to 6, characterized in that, It comprises a power prediction module and a data acquisition module connected thereto; The data acquisition module is used to collect and obtain the location characteristic information of the target power station; wherein the location characteristic information includes the location, altitude, angle and surrounding environment data of each photovoltaic panel; The power prediction module is used to model the target power station based on the location characteristic information, simulate the solar movement trajectory to determine the irradiance time series data of each photovoltaic panel in the target power station; Detect the coverage area of the micro-meteorological phenomenon in the area where the target power station is located, marked as weather area one; predict and obtain the weather prediction data of the target power station and the micro-meteorological data of the weather area one; based on the weather prediction data and the micro-meteorological data, construct the weather time series data of each photovoltaic panel; The detection of the coverage area of the micro-meteorological phenomenon in the area where the target power station is located includes: Extract the historical weather data of the target power station; classify the historical weather data according to meteorological elements to obtain each element data; wherein the meteorological elements include temperature, wind speed, humidity and solar irradiance; Determine whether there is a micro-meteorological phenomenon based on the change trend of each element data; if yes, determine the coverage area of the micro-meteorological phenomenon according to the change trend; if not, no coverage area needs to be determined; The determination of whether there is a micro-meteorological phenomenon based on the change trend of each element data includes: Based on the element data, establish a three-dimensional change graph of the corresponding meteorological element, and identify the abnormal area in the three-dimensional change graph respectively; wherein the three-dimensional change graph is used to represent the change of the meteorological element at different positions. When the abnormal areas corresponding to the at least one meteorological element overlap, it is determined that the area where the target power station is located has a microclimate phenomenon; The determining the coverage area of the microclimate phenomenon according to the change trend comprises: selecting at least one target element from the meteorological elements according to an influence weight on photovoltaic power generation power; wherein the target elements are selected in descending order of the influence weight; taking the abnormal area of the target element as the coverage area of the microclimate phenomenon, or taking the combination of the abnormal areas corresponding to several target elements as the coverage area of the microclimate phenomenon; and, establishing environment time series data of each photovoltaic panel based on the irradiance time series data and the meteorological time series data, inputting the environment time series data into a power prediction model to obtain the power generation power of the target power station; wherein the power prediction model is constructed based on an artificial intelligence model.

Images