A photovoltaic power generation power prediction method

Abstract

The present invention provides a photovoltaic power generation power prediction method, which relates to the technical field of power prediction. The method comprises: acquiring power generation power data of a photovoltaic power generation system and determining equipment impact data; acquiring power generation environment data to construct a photovoltaic power generation scenario and determine environmental impact data; screening and processing the equipment impact data and the environmental impact data based on photovoltaic power generation power prediction settings, constructing a photovoltaic power generation power prediction model to perform a primary independent prediction and a secondary comprehensive prediction, superimposing the prediction results of the primary independent prediction and the prediction results of the secondary comprehensive prediction, and finally outputting the photovoltaic power generation power prediction result; performing application management and regularly maintaining and updating the photovoltaic power generation power prediction model; and effectively achieving efficient and high-quality prediction of photovoltaic power generation.

Description

Technical Field

[0001] The present invention relates to the technical field of power prediction, and in particular to a photovoltaic power generation power prediction method. Background Art

[0002] Green energy has the characteristics of being renewable and having little impact on the environment, and has developed rapidly in recent years. Photovoltaic power generation technology is an effective technology for utilizing green energy, and the scale of photovoltaic power generation is constantly expanding. However, the power of photovoltaic power generation is limited by environmental factors and has periodic intermittent fluctuations. When the power grid is in operation, the power of each power system must be stable or controllable. In addition, the factors that affect the power of photovoltaic power generation are numerous and complex, and the prediction quality of photovoltaic power generation is poor and the prediction efficiency is low.

[0003] Therefore, the present invention provides a photovoltaic power prediction method. Summary of the Invention

[0004] The present invention provides a photovoltaic power generation power prediction method, which is used to determine equipment impact data by acquiring power generation data of a photovoltaic power generation system; acquire power generation environment data to construct a photovoltaic power generation scenario and determine environmental impact data; screen and process the equipment impact data and environmental impact data based on photovoltaic power generation power prediction settings, construct a photovoltaic power generation power prediction model to perform a primary independent prediction and a secondary comprehensive prediction, superimpose the prediction results of the primary independent prediction and the prediction results of the secondary comprehensive prediction, and finally output the photovoltaic power generation power prediction result; perform application management and regularly maintain and update the photovoltaic power generation power prediction model; and effectively achieve efficient and high-quality prediction of photovoltaic power generation.

[0005] The present invention provides a photovoltaic power generation power prediction method, comprising:

[0006] Step 1: Obtain the power generation data of the photovoltaic power generation system and process the data to determine the impact of the equipment type and equipment operating characteristics on the power generation data;

[0007] Step 2: Obtain and process the power generation environment data of the photovoltaic power generation system, construct a photovoltaic power generation scenario, and determine the environmental impact data that affects the power generation data;

[0008] Step 3: Based on the photovoltaic power generation prediction settings, the equipment impact data and environmental impact data are screened and processed to build a photovoltaic power generation prediction model. The photovoltaic power generation prediction settings are input to perform an independent prediction and a secondary comprehensive prediction. The prediction results of the independent prediction and the secondary comprehensive prediction are superimposed to finally output the photovoltaic power generation prediction result.

[0009] Step 4: Based on the photovoltaic power generation power prediction results, the photovoltaic power generation system is managed and the photovoltaic power generation power prediction model is regularly maintained and updated.

[0010] According to a photovoltaic power generation prediction method provided by the present invention, determining the device type impact data and the device operating characteristic impact data that affect the power generation data includes:

[0011] Pre-deploy power generation sensors to monitor and obtain power generation data of photovoltaic power generation systems in real time;

[0012] Unify the format of power generation data, eliminate erroneous power generation data, perform data integration and normalization processing on power generation data based on time series and spatial series, and determine the impact of equipment type and equipment working characteristics on power generation data.

[0013] According to a photovoltaic power generation prediction method provided by the present invention, power generation environment data of a photovoltaic power generation system is obtained and processed, a photovoltaic power generation scenario is constructed, and environmental impact data affecting the power generation data is determined, including:

[0014] Acquire the power generation environment data of the photovoltaic power generation system and perform data cleaning and processing. Filter and integrate the power generation environment data based on time series and spatial series to construct a photovoltaic power generation scenario. The power generation environment data includes: power generation environment temperature data, power generation environment humidity data, and power generation environment wind speed data.

[0015] Conduct scenario-independent and data-holistic analysis on the power generation environment data corresponding to the photovoltaic power generation scenario to determine the scenario data characteristics of the power generation environment data in the same photovoltaic power generation scenario;

[0016] Based on the time proportion of the current photovoltaic power generation scene in all photovoltaic power generation scenes, the same power generation environment data in different photovoltaic power generation scenes are subjected to feature weighted superposition quantification. When the monitored feature weighted superposition quantification value is not lower than the preset impact value, the corresponding power generation environment data is determined to be environmental impact data that affects the power generation power data.

[0017] According to a photovoltaic power generation prediction method provided by the present invention, device impact data and environmental impact data are screened and processed based on photovoltaic power generation prediction settings, including:

[0018] Based on the time series and the spatial series combined with the photovoltaic power generation prediction demand, the photovoltaic power generation prediction setting is obtained, wherein the photovoltaic power generation prediction setting includes: photovoltaic power generation prediction accuracy setting, photovoltaic power generation prediction time setting, photovoltaic power generation prediction space setting, and photovoltaic power generation prediction power setting;

[0019] The equipment impact data and environmental impact data are screened and processed based on the photovoltaic power generation power prediction settings. When there is no correlation between some equipment impact data or some environmental impact data and the photovoltaic power generation power prediction settings, the said part of the equipment impact data or the said part of the environmental impact data are screened and eliminated.

[0020] According to a photovoltaic power generation prediction method provided by the present invention, a photovoltaic power generation prediction setting is obtained based on a time series and a spatial series combined with a photovoltaic power generation prediction demand, including:

[0021] Divide the time series, combine it with the photovoltaic power generation power forecast demand, and obtain the photovoltaic power generation power forecast time setting, wherein the photovoltaic power generation power forecast time setting includes: short-term forecast time setting, medium-term forecast time setting, and long-term forecast time setting;

[0022] The photovoltaic power generation system is divided into spatial sequences according to the photovoltaic power generation units, and the photovoltaic power generation power prediction space setting is obtained in combination with the photovoltaic power generation power prediction demand. The photovoltaic power generation power prediction space setting includes: a prediction small sequence space setting, a prediction medium sequence space setting, and a prediction large sequence space setting.

[0023] The photovoltaic power generation power is obtained based on the user's electricity consumption demand to predict the power demand.

[0024] According to a photovoltaic power generation prediction method provided by the present invention, a photovoltaic power generation prediction model is constructed, photovoltaic power generation prediction settings are input to perform a first independent prediction and a second comprehensive prediction, the prediction results of the first independent prediction and the prediction results of the second comprehensive prediction are superimposed, and finally the photovoltaic power generation power prediction result is output, including:

[0025] Obtain historical photovoltaic power generation prediction data and historical photovoltaic power generation data, and build and train a photovoltaic power generation prediction model;

[0026] Optimizing the parameters of the photovoltaic power generation prediction model based on the equipment impact data and the environmental impact data obtained through screening;

[0027] Perform an independent prediction on the photovoltaic power generation prediction model after optimizing the photovoltaic power generation prediction setting input parameters;

[0028] According to the photovoltaic power generation prediction settings, historical photovoltaic power generation data are retrieved and classified to obtain the historical photovoltaic power generation data set of the predicted time, the historical photovoltaic power generation data set of the predicted space, and the historical photovoltaic power generation data set of the predicted power;

[0029] Sort the data sets according to their data volumes, input the data set with the smallest data volume into the photovoltaic power generation prediction model for an independent prediction, and obtain an independent prediction result of the data set with the smallest data volume;

[0030] Inputting a data set with a medium data volume into the photovoltaic power generation prediction model to perform an independent prediction, obtaining an independent prediction result of the data set with a medium data volume, and superimposing and correcting the independent prediction result of the data set with the smallest data volume;

[0031] Inputting the data set with the largest data volume into the photovoltaic power generation prediction model to perform an independent prediction, obtaining the independent prediction result of the data set with the largest data volume, performing a secondary superposition correction on the independent prediction result of the data set with the smallest data volume after superposition correction, and determining the final independent prediction result;

[0032] Calibrate the photovoltaic power generation power main prediction setting according to the photovoltaic power generation power prediction setting, and obtain the photovoltaic power generation power prediction main device and the photovoltaic power generation power prediction auxiliary device corresponding to the photovoltaic power generation power main prediction setting;

[0033] Screening the photovoltaic power generation prediction auxiliary equipment according to the photovoltaic power generation prediction main equipment, and inputting the historical photovoltaic power generation data of the equipment corresponding to the equipment screening results into the photovoltaic power generation prediction model for secondary comprehensive prediction;

[0034] If the photovoltaic power generation prediction setting weights are consistent, the photovoltaic power generation power main prediction setting is not calibrated, and the photovoltaic power generation equipment with the highest correlation with the photovoltaic power generation power prediction setting is calibrated as the photovoltaic power generation power main prediction equipment;

[0035] Obtaining device parameters of a photovoltaic power generation prediction main device, and determining that the corresponding other photovoltaic power generation devices are photovoltaic power generation auxiliary devices when a parameter similarity value between the device parameters of the other photovoltaic power generation devices and the device parameters of the photovoltaic power generation prediction main device is not less than a preset similarity value;

[0036] Calibrate a standard photovoltaic power generation prediction main device and a standard photovoltaic power generation prediction auxiliary device, and input historical photovoltaic power generation data corresponding to the photovoltaic power generation prediction main device and the historical photovoltaic power generation data corresponding to the photovoltaic power generation prediction auxiliary device into the photovoltaic power generation prediction model to perform a secondary comprehensive prediction;

[0037] Determining the prediction period of the photovoltaic power generation prediction model after parameter optimization according to the photovoltaic power generation prediction accuracy setting, superimposing the final independent prediction and the secondary comprehensive prediction results within the same prediction period to determine the final output photovoltaic power generation prediction value;

[0038] ;in, Indicates the predicted value of the final output photovoltaic power generation; Indicates the last independent prediction input Historical photovoltaic power generation power forecast value; Indicates the The actual value of historical photovoltaic power generation corresponding to the historical photovoltaic power generation power prediction value; Indicates the main equipment for standard photovoltaic power generation prediction; Indicates standard photovoltaic power generation power prediction auxiliary equipment; The first A photovoltaic power generation prediction device; Indicates the parameter similarity value between power generation prediction devices; Indicates the second comprehensive output input The historical photovoltaic power generation prediction value of each photovoltaic power generation prediction device; Indicates the influence weight of an independent prediction on the final output photovoltaic power prediction value; Represents the influence weight of the secondary comprehensive prediction on the final output photovoltaic power prediction value; Indicates the number of historical PV power forecast values ​​input for an independent forecast; Indicates the number of photovoltaic power prediction devices that are input into the secondary comprehensive prediction.

[0039] According to a photovoltaic power generation power prediction method provided by the present invention, application management of a photovoltaic power generation system is performed based on the photovoltaic power generation power prediction result, and a photovoltaic power generation power prediction model is regularly maintained and updated, including:

[0040] Apply management to photovoltaic power generation systems based on photovoltaic power generation power prediction results;

[0041] When the deviation between the photovoltaic power generation prediction result and the actual photovoltaic power generation result is not less than the preset prediction deviation, it is determined that there is a prediction error in the photovoltaic power generation prediction model and a prediction alarm is issued;

[0042] The photovoltaic power generation prediction model is regularly maintained and updated according to the prediction cycle of the photovoltaic power generation prediction model optimized according to the parameter setting determined by the photovoltaic power generation prediction accuracy.

[0043] A photovoltaic power generation power prediction method provided by the present invention further includes:

[0044] Set standard prediction parameters for the photovoltaic power generation prediction model. When the photovoltaic power generation prediction model switches model parameters or there is a prediction error, support directly applying the standard prediction parameters to calibrate the photovoltaic power generation model.

[0045] Visualize the photovoltaic power generation prediction results.

[0046] Compared with the prior art, the present invention has the following advantages:

[0047] Acquire the power generation data of the photovoltaic power generation system and determine the equipment impact data; acquire the power generation environment data to construct the photovoltaic power generation scenario and determine the environmental impact data; screen and process the equipment impact data and environmental impact data based on the photovoltaic power generation power prediction settings, construct a photovoltaic power generation power prediction model to perform an independent prediction and a second comprehensive prediction, superimpose the prediction results of the first independent prediction and the second comprehensive prediction, and finally output the photovoltaic power generation power prediction result; perform application management and regularly maintain and update the photovoltaic power generation power prediction model; effectively achieve efficient and high-quality prediction of photovoltaic power generation.

[0048] Other features and advantages of the present invention will be described in the following description, and in part will become apparent from the description, or will be understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the written description and the accompanying drawings.

[0049] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0051] Figure 1 The present invention provides a flow chart of a photovoltaic power generation prediction method. DETAILED DESCRIPTION

[0052] The preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0053] Example 1:

[0054] The embodiment of the present invention provides a photovoltaic power generation power prediction method, such as Figure 1 Shown, including:

[0055] Step 1: Obtain the power generation data of the photovoltaic power generation system and process the data to determine the impact of the equipment type and equipment operating characteristics on the power generation data;

[0056] Step 2: Obtain and process the power generation environment data of the photovoltaic power generation system, construct a photovoltaic power generation scenario, and determine the environmental impact data that affects the power generation data;

[0057] Step 3: Based on the photovoltaic power generation prediction settings, the equipment impact data and environmental impact data are screened and processed to build a photovoltaic power generation prediction model. The photovoltaic power generation prediction settings are input to perform an independent prediction and a secondary comprehensive prediction. The prediction results of the independent prediction and the secondary comprehensive prediction are superimposed to finally output the photovoltaic power generation prediction result.

[0058] Step 4: Based on the photovoltaic power generation power prediction results, the photovoltaic power generation system is managed and the photovoltaic power generation power prediction model is regularly maintained and updated.

[0059] In this embodiment, the device type that affects the generated power data affects the data, for example, device type a1, generated power b1; device type a2, generated power b2.

[0060] In this embodiment, the equipment operating characteristics affecting the power generation data include, for example, equipment operating state characteristics affecting the data, equipment operating temperature characteristics affecting the data, and equipment operating aging characteristics affecting the data.

[0061] In this embodiment, the photovoltaic power generation power of the photovoltaic power generation system is different in different photovoltaic power generation scenarios.

[0062] In this embodiment, the photovoltaic power generation power prediction setting refers to a spatial prediction setting, a temporal prediction setting, and a power prediction setting.

[0063] In this embodiment, an independent prediction is performed independently according to the photovoltaic power generation power prediction setting.

[0064] In this embodiment, the secondary comprehensive prediction refers to directly inputting the historical photovoltaic power generation data of the photovoltaic power generation equipment into the photovoltaic power generation power prediction model for prediction.

[0065] The working principle and beneficial effects of the above technical solution are: obtaining the power generation data of the photovoltaic power generation system and determining the equipment impact data; obtaining the power generation environment data to construct the photovoltaic power generation scenario and determine the environmental impact data; screening and processing the equipment impact data and the environmental impact data based on the photovoltaic power generation power prediction setting, constructing a photovoltaic power generation power prediction model to perform an independent prediction and a second comprehensive prediction, superimposing the prediction results of the first independent prediction and the prediction results of the second comprehensive prediction, and finally outputting the photovoltaic power generation power prediction result; performing application management and regularly maintaining and updating the photovoltaic power generation power prediction model; effectively realizing efficient and high-quality prediction of photovoltaic power generation.

[0066] Example 2:

[0067] An embodiment of the present invention provides a photovoltaic power generation prediction method, which determines device type impact data and device operating characteristic impact data that affect power generation data, including:

[0068] Pre-deploy power generation sensors to monitor and obtain power generation data of photovoltaic power generation systems in real time;

[0069] Unify the format of power generation data, eliminate erroneous power generation data, perform data integration and normalization processing on power generation data based on time series and spatial series, and determine the impact of equipment type and equipment working characteristics on power generation data.

[0070] In this embodiment, data integration and normalization processing is performed on the power generation data based on the time series and the space series to ensure that the photovoltaic power generation data in the same time series or the same space series are at the same data level, that is, the impact level is consistent.

[0071] The working principle and beneficial effect of the above technical solution are: by determining the equipment type impact data and equipment working characteristics impact data that affect the power generation data, the equipment impact data foundation is laid for subsequent photovoltaic power generation power prediction.

[0072] Example 3:

[0073] An embodiment of the present invention provides a photovoltaic power generation prediction method, which obtains power generation environment data of a photovoltaic power generation system and processes the data, constructs a photovoltaic power generation scenario, and determines environmental impact data that affects the power generation data, including:

[0074] Acquire the power generation environment data of the photovoltaic power generation system and perform data cleaning and processing. Filter and integrate the power generation environment data based on time series and spatial series to construct a photovoltaic power generation scenario. The power generation environment data includes: power generation environment temperature data, power generation environment humidity data, and power generation environment wind speed data.

[0075] Conduct scenario-independent and data-holistic analysis on the power generation environment data corresponding to the photovoltaic power generation scenario to determine the scenario data characteristics of the power generation environment data in the same photovoltaic power generation scenario;

[0076] Based on the time proportion of the current photovoltaic power generation scene in all photovoltaic power generation scenes, the same power generation environment data in different photovoltaic power generation scenes are subjected to feature weighted superposition quantification. When the monitored feature weighted superposition quantification value is not lower than the preset impact value, the corresponding power generation environment data is determined to be environmental impact data that affects the power generation power data.

[0077] In this embodiment, performing scene-independent and data-integrated analysis on power generation environment data corresponding to a photovoltaic power generation scene refers to performing an overall analysis on the power generation environment data in the same scene.

[0078] In this embodiment, feature weighted superposition and quantification are performed on the same power generation environment data in different photovoltaic power generation scenarios based on the time proportion of the current photovoltaic power generation scenario in all photovoltaic power generation scenarios. For example, the key power generation data a1 exists in the current photovoltaic power generation scenarios b1 and b2, the current photovoltaic power generation scenario b1 accounts for 0.2 of the time in all photovoltaic power generation scenarios, and the current photovoltaic power generation scenario b1 accounts for 0.4 of the time in all photovoltaic power generation scenarios. Feature weighted superposition and quantification are performed based on the time proportion.

[0079] The working principle and beneficial effects of the above technical solution are: obtaining the power generation environment data of the photovoltaic power generation system and performing data processing, constructing a photovoltaic power generation scenario, determining the environmental impact data that affects the power generation data, and laying the environmental impact data foundation for subsequent photovoltaic power generation power prediction.

[0080] Example 4:

[0081] An embodiment of the present invention provides a photovoltaic power generation prediction method, which filters and processes equipment impact data and environmental impact data based on photovoltaic power generation prediction settings, including:

[0082] Based on the time series and the spatial series combined with the photovoltaic power generation prediction demand, the photovoltaic power generation prediction setting is obtained, wherein the photovoltaic power generation prediction setting includes: photovoltaic power generation prediction accuracy setting, photovoltaic power generation prediction time setting, photovoltaic power generation prediction space setting, and photovoltaic power generation prediction power setting;

[0083] The equipment impact data and environmental impact data are screened and processed based on the photovoltaic power generation power prediction settings. When there is no correlation between some equipment impact data or some environmental impact data and the photovoltaic power generation power prediction settings, the said part of the equipment impact data or the said part of the environmental impact data are screened and eliminated.

[0084] In this embodiment, the photovoltaic power generation power prediction accuracy is set, that is, the prediction error is controlled. For example, if the photovoltaic power generation power prediction accuracy is set to 0.9, the prediction error cannot exceed 0.1.

[0085] In this embodiment, the photovoltaic power generation power prediction time is set, for example, the photovoltaic power generation power in the time period t1 is predicted.

[0086] In this embodiment, the photovoltaic power generation prediction space is set, for example, the photovoltaic power generation in the area a1 is set to be predicted.

[0087] In this embodiment, the photovoltaic power generation power prediction power setting, for example, performs power prediction on the photovoltaic power generation equipment of b1 photovoltaic power generation power.

[0088] In this embodiment, when there is no correlation between the equipment impact data or the environmental impact data and the photovoltaic power generation power prediction setting, for example, when setting the prediction of the photovoltaic power generation power of the photovoltaic power generation equipment of type d1 in the c1 area and the t2 time period, the environmental impact data of the c2 area is eliminated.

[0089] The working principle and beneficial effects of the above technical solution are: based on the photovoltaic power generation prediction setting, the equipment impact data and environmental impact data are screened and processed, the impact data is simplified, which is conducive to improving the prediction efficiency of photovoltaic power generation.

[0090] Example 5:

[0091] An embodiment of the present invention provides a photovoltaic power generation power prediction method, which obtains photovoltaic power generation power prediction settings based on time series and spatial series combined with photovoltaic power generation power prediction requirements, including:

[0092] Divide the time series, combine it with the photovoltaic power generation power forecast demand, and obtain the photovoltaic power generation power forecast time setting, wherein the photovoltaic power generation power forecast time setting includes: short-term forecast time setting, medium-term forecast time setting, and long-term forecast time setting;

[0093] The photovoltaic power generation system is divided into spatial sequences according to the photovoltaic power generation units, and the photovoltaic power generation power prediction space setting is obtained in combination with the photovoltaic power generation power prediction demand. The photovoltaic power generation power prediction space setting includes: a prediction small sequence space setting, a prediction medium sequence space setting, and a prediction large sequence space setting.

[0094] The photovoltaic power generation power is obtained based on the user's electricity consumption demand to predict the power demand.

[0095] In this embodiment, the prediction accuracy and prediction efficiency corresponding to the short-term prediction time setting, the mid-term prediction time setting, and the long-term prediction time setting are different.

[0096] In this embodiment, the small, medium and large sequence space settings can be set according to the photovoltaic power generation equipment or the power consumption area.

[0097] In this embodiment, the photovoltaic power generation power forecast demand is obtained according to the user's power demand. For example, if the user's power demand changes to a1 during the period t1, the power generation power of each photovoltaic power generation device during the period t1 needs to be predicted.

[0098] The working principle and beneficial effects of the above technical solution are: based on the time series and spatial series combined with the photovoltaic power generation power prediction needs, the photovoltaic power generation power prediction settings are obtained, which facilitates the subsequent effective prediction of photovoltaic power generation and improves the prediction quality.

[0099] Example 6:

[0100] An embodiment of the present invention provides a photovoltaic power generation prediction method, which constructs a photovoltaic power generation prediction model, inputs photovoltaic power generation prediction settings, performs a primary independent prediction and a secondary comprehensive prediction, superimposes the prediction results of the primary independent prediction and the prediction results of the secondary comprehensive prediction, and finally outputs the photovoltaic power generation prediction result, including:

[0101] Obtain historical photovoltaic power generation prediction data and historical photovoltaic power generation data, and build and train a photovoltaic power generation prediction model;

[0102] Optimizing the parameters of the photovoltaic power generation prediction model based on the equipment impact data and the environmental impact data obtained through screening;

[0103] Perform an independent prediction on the photovoltaic power generation prediction model after optimizing the photovoltaic power generation prediction setting input parameters;

[0104] According to the photovoltaic power generation prediction settings, historical photovoltaic power generation data are retrieved and classified to obtain the historical photovoltaic power generation data set of the predicted time, the historical photovoltaic power generation data set of the predicted space, and the historical photovoltaic power generation data set of the predicted power;

[0105] Sort the data sets according to their data volumes, input the data set with the smallest data volume into the photovoltaic power generation prediction model for an independent prediction, and obtain an independent prediction result of the data set with the smallest data volume;

[0106] Inputting a data set with a medium data volume into the photovoltaic power generation prediction model to perform an independent prediction, obtaining an independent prediction result of the data set with a medium data volume, and superimposing and correcting the independent prediction result of the data set with the smallest data volume;

[0107] Inputting the data set with the largest data volume into the photovoltaic power generation prediction model to perform an independent prediction, obtaining the independent prediction result of the data set with the largest data volume, performing a secondary superposition correction on the independent prediction result of the data set with the smallest data volume after superposition correction, and determining the final independent prediction result;

[0108] Calibrate the photovoltaic power generation power main prediction setting according to the photovoltaic power generation power prediction setting, and obtain the photovoltaic power generation power prediction main device and the photovoltaic power generation power prediction auxiliary device corresponding to the photovoltaic power generation power main prediction setting;

[0109] Screening the photovoltaic power generation prediction auxiliary equipment according to the photovoltaic power generation prediction main equipment, and inputting the historical photovoltaic power generation data of the equipment corresponding to the equipment screening results into the photovoltaic power generation prediction model for secondary comprehensive prediction;

[0110] If the photovoltaic power generation prediction setting weights are consistent, the photovoltaic power generation power main prediction setting is not calibrated, and the photovoltaic power generation equipment with the highest correlation with the photovoltaic power generation power prediction setting is calibrated as the photovoltaic power generation power main prediction equipment;

[0111] Obtaining device parameters of a photovoltaic power generation prediction main device, and determining that the corresponding other photovoltaic power generation devices are photovoltaic power generation auxiliary devices when a parameter similarity value between the device parameters of the other photovoltaic power generation devices and the device parameters of the photovoltaic power generation prediction main device is not less than a preset similarity value;

[0112] Calibrate a standard photovoltaic power generation prediction main device and a standard photovoltaic power generation prediction auxiliary device, and input historical photovoltaic power generation data corresponding to the photovoltaic power generation prediction main device and the historical photovoltaic power generation data corresponding to the photovoltaic power generation prediction auxiliary device into the photovoltaic power generation prediction model to perform a secondary comprehensive prediction;

[0113] Determining the prediction period of the photovoltaic power generation prediction model after parameter optimization according to the photovoltaic power generation prediction accuracy setting, superimposing the final independent prediction and the secondary comprehensive prediction results within the same prediction period to determine the final output photovoltaic power generation prediction value;

[0114] ;in, Indicates the predicted value of the final output photovoltaic power generation; Indicates the last independent prediction input Historical photovoltaic power generation power forecast value; Indicates the The actual value of historical photovoltaic power generation corresponding to the historical photovoltaic power generation power prediction value; Indicates the main equipment for standard photovoltaic power generation prediction; Indicates standard photovoltaic power generation power prediction auxiliary equipment; The first A photovoltaic power generation prediction device; Indicates the parameter similarity value between power generation prediction devices; Indicates the second comprehensive output input The historical photovoltaic power generation prediction value of each photovoltaic power generation prediction device; Indicates the influence weight of an independent prediction on the final output photovoltaic power prediction value; Represents the influence weight of the secondary comprehensive prediction on the final output photovoltaic power prediction value; Indicates the number of historical PV power forecast values ​​input for an independent forecast; Indicates the number of photovoltaic power prediction devices that are input into the secondary comprehensive prediction.

[0115] In this embodiment, sorting is performed according to the data volume of the data set. For example, the data volume a1 of the predicted time historical photovoltaic power generation dataset, the data volume a1 of the predicted space historical photovoltaic power generation dataset, and the data volume a1 of the predicted power historical photovoltaic power generation dataset, where a1 < a2 < a3. First, the predicted time historical photovoltaic power generation dataset is input into the photovoltaic power generation prediction model for an independent prediction, then the predicted space historical photovoltaic power generation dataset is input for an independent prediction, and finally the predicted power historical photovoltaic power generation dataset is input for an independent prediction.

[0116] In this embodiment, an independent prediction result of the medium data volume dataset is obtained, and the independent prediction result of the smallest data volume dataset is superimposed and corrected. For example, the independent prediction result of the smallest data volume dataset defines a time period t1, and the independent prediction result of the medium data volume dataset defines a spatial region b1.

[0117] In this embodiment, an independent prediction result of the largest data volume dataset is obtained, and the independent prediction result of the smallest data volume dataset after superimposed correction is subjected to a secondary superimposed correction to determine the final independent prediction result. For example, for the secondary superimposed correction, a power range c1 is defined, and the final independent prediction result is the photovoltaic power generation power for the predicted time period t1, spatial region b1, and power range c1.

[0118] In this embodiment, the main device for photovoltaic power generation prediction refers to the main prediction device for performing photovoltaic power generation prediction.

[0119] In this embodiment, the auxiliary device for photovoltaic power generation prediction refers to the secondary prediction device for performing photovoltaic power generation prediction.

[0120] In this embodiment, device screening is performed on the auxiliary device for photovoltaic power generation prediction according to the main device for photovoltaic power generation prediction. For example, device screening can be performed according to the parameter similarity value between the main device for photovoltaic power generation prediction and the auxiliary device for photovoltaic power generation prediction.

[0121] In this embodiment, the higher the parameter similarity value between the main device for photovoltaic power generation prediction and the auxiliary device for photovoltaic power generation prediction, the simpler the photovoltaic power generation prediction, and thus the quality of photovoltaic power generation prediction is improved.

[0122] The working principle and beneficial effects of the above technical solution are as follows: By constructing a photovoltaic power generation prediction model, inputting the photovoltaic power generation prediction settings for an independent prediction and a secondary comprehensive prediction, superimposing the prediction results of the independent prediction and the secondary comprehensive prediction, and finally outputting the photovoltaic power generation prediction result, the efficient and high-quality prediction of photovoltaic power generation is effectively achieved.

[0123] Example 7:

[0124] An embodiment of the present invention provides a photovoltaic power generation power prediction method, which manages the photovoltaic power generation system based on the photovoltaic power generation power prediction result and regularly maintains and updates the photovoltaic power generation power prediction model, including:

[0125] Apply management to photovoltaic power generation systems based on photovoltaic power generation power prediction results;

[0126] When the deviation between the photovoltaic power generation prediction result and the actual photovoltaic power generation result is not less than the preset prediction deviation, it is determined that there is a prediction error in the photovoltaic power generation prediction model and a prediction alarm is issued;

[0127] The photovoltaic power generation prediction model is regularly maintained and updated according to the prediction cycle of the photovoltaic power generation prediction model optimized according to the parameter setting determined by the photovoltaic power generation prediction accuracy.

[0128] In this embodiment, application management of the photovoltaic power generation system is performed based on the photovoltaic power generation power prediction result, for example, some photovoltaic power generation equipment is disabled to ensure stable operation of the power grid.

[0129] In this embodiment, the higher the photovoltaic power generation power prediction accuracy is set, the shorter the prediction period and the maintenance update period are.

[0130] The working principle and beneficial effects of the above technical solution are: based on the photovoltaic power generation power prediction results, the photovoltaic power generation system is applied and managed, and the photovoltaic power generation power prediction model is regularly maintained and updated, ensuring the long-term prediction stability and effectiveness of photovoltaic power generation.

[0131] Example 8:

[0132] An embodiment of the present invention provides a photovoltaic power generation power prediction method, further comprising:

[0133] Set standard prediction parameters for the photovoltaic power generation prediction model. When the photovoltaic power generation prediction model switches model parameters or there is a prediction error, support directly applying the standard prediction parameters to calibrate the photovoltaic power generation model.

[0134] Visualize the photovoltaic power generation prediction results.

[0135] The standard prediction parameters are applied to calibrate the standard prediction parameters of the photovoltaic power generation model. For example, when predicting the photovoltaic power generation power in the a1 area during the t1 period, the b1 parameter of the photovoltaic power generation power prediction model is adjusted; when switching to predicting the photovoltaic power generation power in the a1 area during the t2 period, the standard prediction parameters are applied first, and the b1 parameter is adjusted to the standard prediction parameter, and then adjusted to the c2 parameter.

[0136] In this embodiment, different photovoltaic power generation power predictions require different parameter adjustments to the photovoltaic power generation power prediction model.

[0137] The working principle and beneficial effects of the above technical solution are: by visually displaying the photovoltaic power generation power prediction results, it is convenient to perform subsequent operations such as power control; by setting standard prediction parameters, the efficiency of photovoltaic power generation power prediction is further improved and the prediction quality is ensured.

[0138] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A photovoltaic power generation power prediction method, characterized in that: include: Step 1: Obtain the power generation data of the photovoltaic power generation system and process the data to determine the impact of the equipment type and equipment operating characteristics on the power generation data; Step 2: Obtain and process the power generation environment data of the photovoltaic power generation system, construct a photovoltaic power generation scenario, and determine the environmental impact data that affects the power generation data; Step 3: Based on the photovoltaic power generation prediction settings, the equipment impact data and environmental impact data are screened and processed to build a photovoltaic power generation prediction model. The photovoltaic power generation prediction settings are input to perform an independent prediction and a secondary comprehensive prediction. The prediction results of the independent prediction and the secondary comprehensive prediction are superimposed to finally output the photovoltaic power generation prediction result. Among them, an independent prediction first retrieves historical photovoltaic power generation data according to the photovoltaic power generation prediction setting classification to obtain the prediction time historical photovoltaic power generation data set, the prediction space historical photovoltaic power generation data set and the prediction power historical photovoltaic power generation data set; the data sets are sorted according to the size of the data set, and the data set with the smallest data set is first input into the model for an independent prediction to obtain the first prediction result, and then the data set with medium data set is used for prediction to obtain the second prediction result, the first prediction result is superimposed and corrected, and then the data set with the largest data set is used for prediction to obtain the third prediction result, and the corrected first prediction result is superimposed and corrected again to determine the final independent prediction result; The secondary comprehensive prediction calibrates the prediction main device and the prediction auxiliary device according to the prediction setting and the main prediction setting, or determines the prediction main device and the prediction auxiliary device through correlation and device parameter similarity, and inputs the historical photovoltaic power generation data corresponding to the photovoltaic power generation prediction main device and the historical photovoltaic power generation data of the photovoltaic power generation auxiliary device into the photovoltaic power generation prediction model for secondary comprehensive prediction; Step 4: Based on the photovoltaic power generation power prediction results, the photovoltaic power generation system is managed and the photovoltaic power generation power prediction model is regularly maintained and updated.

2. A photovoltaic power generation power prediction method according to claim 1, characterized in that: Determine the impact of equipment type and equipment operating characteristics on power generation data, including: Pre-deploy power generation sensors to monitor and obtain power generation data of photovoltaic power generation systems in real time; Unify the format of power generation data, eliminate erroneous power generation data, perform data integration and normalization processing on power generation data based on time series and spatial series, and determine the impact of equipment type and equipment working characteristics on power generation data.

3. A photovoltaic power generation power prediction method according to claim 1, characterized in that: Acquire and process the power generation environment data of the photovoltaic power generation system, construct photovoltaic power generation scenarios, and determine the environmental impact data that affects the power generation data, including: Acquire the power generation environment data of the photovoltaic power generation system and perform data cleaning and processing. Filter and integrate the power generation environment data based on time series and spatial series to construct a photovoltaic power generation scenario. The power generation environment data includes: power generation environment temperature data, power generation environment humidity data, and power generation environment wind speed data. Conduct scenario-independent and data-holistic analysis on the power generation environment data corresponding to the photovoltaic power generation scenario to determine the scenario data characteristics of the power generation environment data in the same photovoltaic power generation scenario; Based on the time proportion of the current photovoltaic power generation scene in all photovoltaic power generation scenes, the same power generation environment data in different photovoltaic power generation scenes are subjected to feature weighted superposition quantification. When the monitored feature weighted superposition quantification value is not lower than the preset impact value, the corresponding power generation environment data is determined to be environmental impact data that affects the power generation power data.

4. The photovoltaic power generation prediction method according to claim 1, characterized in that: Based on the photovoltaic power generation prediction settings, the equipment impact data and environmental impact data are screened and processed, including: Based on the time series and the spatial series combined with the photovoltaic power generation prediction demand, the photovoltaic power generation prediction setting is obtained, wherein the photovoltaic power generation prediction setting includes: photovoltaic power generation prediction accuracy setting, photovoltaic power generation prediction time setting, photovoltaic power generation prediction space setting, and photovoltaic power generation prediction power setting; The equipment impact data and environmental impact data are screened and processed based on the photovoltaic power generation power prediction settings. When there is no correlation between some equipment impact data or some environmental impact data and the photovoltaic power generation power prediction settings, the said part of the equipment impact data or the said part of the environmental impact data are screened and eliminated.

5. A photovoltaic power generation power prediction method according to claim 4, characterized in that: Based on the time series and spatial series combined with the photovoltaic power generation forecast demand, the photovoltaic power generation forecast settings are obtained, including: Divide the time series, combine it with the photovoltaic power generation power forecast demand, and obtain the photovoltaic power generation power forecast time setting, wherein the photovoltaic power generation power forecast time setting includes: short-term forecast time setting, medium-term forecast time setting, and long-term forecast time setting; Perform spatial sequence division according to the photovoltaic power generation units of the photovoltaic power generation system, and obtain photovoltaic power generation power prediction space settings in combination with photovoltaic power generation power prediction requirements. The photovoltaic power generation power prediction space settings include: prediction small sequence space settings, prediction medium sequence space settings, and prediction large sequence space settings. The photovoltaic power generation power is obtained based on the user's electricity consumption demand to predict the power demand.

6. A photovoltaic power generation power prediction method according to claim 1, characterized in that: Construct a photovoltaic power generation prediction model, input the photovoltaic power generation prediction settings to perform an independent prediction and a secondary comprehensive prediction, superimpose the prediction results of the independent prediction and the secondary comprehensive prediction, and finally output the photovoltaic power generation prediction results, including: Obtain historical photovoltaic power generation prediction data and historical photovoltaic power generation data, and build and train a photovoltaic power generation prediction model; Optimizing the parameters of the photovoltaic power generation prediction model based on the equipment impact data and the environmental impact data obtained through screening; Perform an independent prediction on the photovoltaic power generation prediction model after optimizing the photovoltaic power generation prediction setting input parameters; According to the photovoltaic power generation prediction settings, historical photovoltaic power generation data are retrieved and classified to obtain the historical photovoltaic power generation data set of the predicted time, the historical photovoltaic power generation data set of the predicted space, and the historical photovoltaic power generation data set of the predicted power; Sort the data sets according to their data size, input the data set with the smallest data size into the photovoltaic power generation power prediction model for an independent prediction, and obtain an independent prediction result of the data set with the smallest data size; Inputting a data set with a medium data volume into the photovoltaic power generation prediction model to perform an independent prediction, obtaining an independent prediction result of the data set with a medium data volume, and performing superposition correction on the independent prediction result of the data set with the smallest data volume; Inputting the data set with the largest data volume into the photovoltaic power generation prediction model for an independent prediction, obtaining the independent prediction result of the data set with the largest data volume, performing a secondary superposition correction on the independent prediction result of the data set with the smallest data volume after superposition correction, and determining the final independent prediction result; Calibrate the photovoltaic power generation power main prediction setting according to the photovoltaic power generation power prediction setting, and obtain the photovoltaic power generation power prediction main device and the photovoltaic power generation power prediction auxiliary device corresponding to the photovoltaic power generation power main prediction setting; Screening the photovoltaic power generation prediction auxiliary equipment according to the photovoltaic power generation prediction main equipment, and inputting the historical photovoltaic power generation data of the equipment corresponding to the equipment screening results into the photovoltaic power generation prediction model for secondary comprehensive prediction; If the photovoltaic power generation prediction setting weights are consistent, the photovoltaic power generation power main prediction setting is not calibrated, and the photovoltaic power generation equipment with the highest correlation with the photovoltaic power generation power prediction setting is calibrated as the photovoltaic power generation power main prediction equipment; Obtaining device parameters of a photovoltaic power generation prediction main device, and determining that the corresponding other photovoltaic power generation devices are photovoltaic power generation auxiliary devices when a parameter similarity value between the device parameters of the other photovoltaic power generation devices and the device parameters of the photovoltaic power generation prediction main device is not less than a preset similarity value; Calibrate a standard photovoltaic power generation prediction main device and a standard photovoltaic power generation prediction auxiliary device, and input historical photovoltaic power generation data corresponding to the photovoltaic power generation prediction main device and the historical photovoltaic power generation data of the photovoltaic power generation prediction auxiliary device into the photovoltaic power generation prediction model to perform a secondary comprehensive prediction; Determining the prediction period of the photovoltaic power generation prediction model after parameter optimization according to the photovoltaic power generation prediction accuracy setting, superimposing the final independent prediction and the secondary comprehensive prediction results within the same prediction period to determine the final output photovoltaic power generation prediction value; ;in, Indicates the predicted value of the final output photovoltaic power generation; Indicates the last independent prediction input Historical photovoltaic power generation power forecast value; Indicates the The actual value of historical photovoltaic power generation corresponding to the historical photovoltaic power generation power prediction value; Indicates the main equipment for standard photovoltaic power generation prediction; Indicates standard photovoltaic power generation power prediction auxiliary equipment; The first A photovoltaic power generation prediction device; Indicates the parameter similarity value between power generation prediction devices; Indicates the second comprehensive output input The historical photovoltaic power generation prediction value of each photovoltaic power generation prediction device; Indicates the influence weight of an independent prediction on the final output photovoltaic power prediction value; Represents the influence weight of the secondary comprehensive prediction on the final output photovoltaic power prediction value; Indicates the number of historical PV power forecast values ​​input for an independent forecast; Indicates the number of photovoltaic power prediction devices that are input into the secondary comprehensive prediction.

7. A photovoltaic power generation power prediction method according to claim 1, characterized in that: Based on the photovoltaic power generation power prediction results, the photovoltaic power generation system is managed and the photovoltaic power generation power prediction model is regularly maintained and updated, including: Apply management to photovoltaic power generation systems based on photovoltaic power generation power prediction results; When the deviation between the photovoltaic power generation prediction result and the actual photovoltaic power generation result is not less than the preset prediction deviation, it is determined that there is a prediction error in the photovoltaic power generation prediction model and a prediction alarm is issued; The photovoltaic power generation prediction model is regularly maintained and updated according to the prediction cycle of the photovoltaic power generation prediction model optimized according to the parameter setting determined by the photovoltaic power generation prediction accuracy.

8. The photovoltaic power generation prediction method according to claim 1, characterized in that: Also includes: Set standard prediction parameters for the photovoltaic power generation prediction model. When the photovoltaic power generation prediction model switches model parameters or there is a prediction error, support directly applying the standard prediction parameters to calibrate the photovoltaic power generation model. Visualize the photovoltaic power generation prediction results.

Images