Real-time calculation method and system for distributed photovoltaic power generation meteorological elements

Abstract

The invention discloses a real-time calculation method and system for distributed photovoltaic power generation meteorological elements, and solves the technical problem of regional distributed photovoltaic power generation meteorological information loss. The method comprises the following steps: selecting adjacent sites based on power station operation data, and repairing target site data by using adjacent site data; constructing an irradiance calculation model by using the reconstructed power data, and establishing a model among numerical weather forecast, distributed photovoltaic station power, irradiance and environment temperature; and based on the established model, calculating the solar irradiance and temperature of the position of the distributed photovoltaic station by taking the real-time power and the current value of the numerical weather forecast as input. In order to complement the power of the distributed photovoltaic station, a meteorological information calculation model is established, a high-cost-performance and high-precision solution is provided for complementing the meteorological information of distributed photovoltaic power generation, and the accuracy of output power prediction of the distributed photovoltaic station is improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of photovoltaic power generation, and in particular relates to a real-time calculation method and system for meteorological elements of distributed photovoltaic power generation. Background Art

[0002] With the transformation of energy structures and the rapid development of renewable energy, distributed photovoltaic power generation has become a key development direction in the global energy sector due to its high resource utilization, low environmental impact, clean and efficient operation, and immediate use. However, for the intelligent operation and maintenance and precise power forecasting of distributed photovoltaic power generation, real-time and accurate measurement of meteorological factors such as solar irradiance and ambient temperature is a key prerequisite. Currently, the acquisition of meteorological information mainly relies on ground-based meteorological stations. However, due to the high construction costs, distributed photovoltaic power stations are generally not equipped with independent meteorological monitoring equipment, resulting in a significant problem of missing meteorological data. In addition, meteorological stations within a region are sparsely distributed and lack spatial representativeness, making it difficult to accurately reflect the micrometeorological conditions at the locations of each distributed photovoltaic power station. Especially in areas with complex terrain or variable climate, the deviation between existing meteorological data and the actual environmental parameters of the power station is further exacerbated, seriously restricting the accuracy of power forecasting and the efficiency of power station operation and maintenance.

[0003] In existing technologies, some solutions attempt to use data from neighboring meteorological stations to complete the missing information of the target site through interpolation or simple weighted averaging. However, such methods do not fully consider the spatial heterogeneity between distributed photovoltaic power stations and the dynamic correlation between meteorological elements, resulting in large errors in the completion results. In addition, although numerical weather forecasts can provide large-scale meteorological forecast data, their spatial resolution is limited, making it difficult to directly match the local meteorological characteristics of distributed photovoltaic power stations. On the other hand, modeling methods based on a single data source often lack the accuracy of meteorological parameter inversion due to incomplete input information, and cannot meet the high real-time and high-precision application requirements.

[0004] Therefore, how to achieve real-time and accurate calculation of solar irradiance and temperature at the location of distributed photovoltaic power stations in a low-cost and high-efficiency manner, break through the bottleneck of missing meteorological information, and improve power prediction and operation and maintenance levels has become a technical problem that needs to be solved urgently in this field. Summary of the Invention

[0005] In response to the above-mentioned deficiencies in the prior art, the present invention proposes a real-time calculation method and system for meteorological elements of distributed photovoltaic power generation, which is of great significance for realizing intelligent operation and safe access of distributed photovoltaic power generation.

[0006] To achieve the above objectives, the present invention provides the following technical solutions:

[0007] In a first aspect, the present invention provides a method for real-time calculation of meteorological elements of distributed photovoltaic power generation, comprising the following steps:

[0008] S1. Selecting neighboring sites based on the operating data of the target power station, and using the data of the neighboring sites to complete the missing power data of the target site and replace abnormal values;

[0009] S2. Using the completed power data and irradiance data from the numerical weather forecast, a neural network-based meteorological element calculation model is constructed to establish a mapping relationship between the numerical weather forecast, the power of distributed photovoltaic sites, and solar irradiance and ambient temperature;

[0010] S3. Using the real-time power data and the current value of the numerical weather forecast as input, the solar irradiance and ambient temperature at the location of the target power station are calculated in real time using the meteorological element calculation model.

[0011] Furthermore, step S1 specifically includes:

[0012] S1-1. Integrate power data from the photovoltaic power station operation data in the region;

[0013] S1-2. Fill in the missing data of the target power station and replace outliers. Select the target point x0, calculate the distance between all power stations and the target power station, and select the 5 closest power stations as adjacent reference points. The distance calculation formula is:

[0014]

[0015] Where r is the radius of the Earth (6378 km), Δlon is the longitude difference between the neighboring site and the target site, Δlat is the latitude difference between the neighboring site and the target site, lat1 is the latitude of the neighboring DPV site, and lat2 is the latitude of the target DPV site;

[0016] S1-3, based on the geographical distance between the adjacent reference site and the target site, calculate the weight w(x0, x i ), the weight is a function of the inverse of the feature distance, and the calculation formula is:

[0017]

[0018] in, f i is the feature vector of the neighboring site, f is the feature vector of the target site, d(x0,x i ) is the characteristic distance between the neighboring site and the target site;

[0019] S1-4, using the measured value y(x i ) Calculate the fitting value of the target power station That is the completed distributed photovoltaic power station power, the calculation formula is:

[0020]

[0021] Furthermore, in step S2, the modeling process for solar radiation data specifically includes the following sub-steps:

[0022] S2-1. Use the completed distributed photovoltaic power station power and irradiance data from numerical weather forecast as input data;

[0023] S2-2, calculate the forget gate output f based on the current input and the hidden state of the previous time step t , determines which information in the current memory unit needs to be forgotten, and its calculation formula is:

[0024] f t =σ(W f [h t-1 ,x t ]+b f );

[0025] Where: σ represents the sigmoid function, W f Represents the forget gate weight, h t-1 represents the hidden state from the previous time step t-1, x t represents the input of the current time step t, b f Represents the forget gate bias;

[0026] S2-3, calculate the output value i of the input gate t and candidate memory cell values And add the results of the forget gate and the input gate to update the memory cell state c of the current time step t :

[0027] i t =σ(W i [h t-1 ,x t ]+b i );

[0028]

[0029] Where: W i Represents the input gate weight, W c represents the candidate unit state weight, c t-1 represents the cell state at the previous time step, b i represents the input gate bias, b c Represents the candidate unit state bias;

[0030] S2-4, according to the current input x t and the memory cell state c t , calculate the output value o of the output gatet , get the new hidden state h t , this state will be passed to the next time step as output and also serve as the output of the current time step:

[0031] o t =σ(W o [h t-1 ,x t ]+b o );

[0032] h t =o t *tanh(c t );

[0033] Where: W o represents the output gate weight, b o represents the output gate bias;

[0034] S2-5. Compare the output irradiance value with the actual irradiance value and calculate whether the deviation d between the two meets the requirements. If d meets the requirements, obtain the specified weights and biases. If d does not meet the requirements, update the weights and biases through the back propagation algorithm and repeat steps S2-2 to S2-4 until the end condition is met.

[0035] S2-6. Obtain calculated irradiation data.

[0036] Furthermore, in step S2, the modeling process for temperature data specifically includes the following sub-steps:

[0037] S3-1, using the completed distributed photovoltaic power station power and temperature data from numerical weather forecasting as input data;

[0038] S3-2, repeat steps S2-2 to S2-4 until the end condition is met;

[0039] S3-3. Obtain calculated temperature data.

[0040] Furthermore, the meteorological element calculation model is a LSTM-based neural network model.

[0041] In a second aspect, the present invention provides a real-time calculation system for meteorological elements of distributed photovoltaic power generation, comprising:

[0042] A data repair module is used to select adjacent sites based on power plant operation data and use the data of adjacent sites to repair the target site data;

[0043] The model building module is used to build a meteorological element calculation model using the repaired power data, and establish a model between numerical weather forecast, distributed photovoltaic site power and irradiance, and ambient temperature;

[0044] The calculation module is used to calculate the solar irradiance and temperature at the location of the distributed photovoltaic site based on the constructed model, taking the real-time power and the current value of the numerical weather forecast as input.

[0045] In a third aspect, the present invention provides an electronic device comprising: a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein the processor implements the steps in a method for real-time calculation of meteorological elements of distributed photovoltaic power generation when executing the computer program.

[0046] In a fourth aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps in the real-time calculation method of meteorological elements of distributed photovoltaic power generation.

[0047] Based on the above technical solution, the embodiments of the present invention can produce at least the following technical effects:

[0048] The present invention provides a real-time calculation method and system for meteorological elements of distributed photovoltaic power generation. This method takes into account the problem that poor data quality of distributed photovoltaic power stations has a significant impact on the results. It reconstructs the power data of power-missing sites, repairs the missing information of the target site by integrating the operating data of neighboring photovoltaic sites, reconstructs the power data by combining numerical weather forecasting and an adaptive weighted fusion algorithm, and uses a neural network model to establish a meteorological element calculation model, effectively solving the problem of missing meteorological information of distributed photovoltaic power stations; significantly improves the accuracy and real-time performance of solar irradiance and temperature calculations, reduces meteorological monitoring costs, and supports accurate power prediction, intelligent operation and maintenance, and secure access of distributed photovoltaic power generation through high-resolution meteorological data, providing a cost-effective and high-precision solution for supplementing meteorological information of distributed photovoltaic power generation, while helping to improve the accuracy of output power prediction of distributed photovoltaic power stations. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the structures shown in these drawings without paying any creative work.

[0050] Figure 1 is a flow chart of the method of the present invention;

[0051] Figure 2 This is a schematic diagram of the present invention when using a neural network for modeling;

[0052] Figure 3The irradiance map calculated based on the meteorological information of each power station in Example 1 of the present invention;

[0053] Figure 4 This is a temperature map calculated based on the meteorological information of each power station in Example 1 of the present invention. DETAILED DESCRIPTION

[0054] The technical solutions in the embodiments of the present invention will be described clearly and completely below. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention. In addition, the technical solutions between the various embodiments can be combined with each other, but they must be based on the ability of ordinary technicians in this field to implement them. When the combination of technical solutions is contradictory or cannot be implemented, it should be deemed that such a combination of technical solutions does not exist and is not within the scope of protection required by the present invention.

[0055] like Figure 1 As shown, the real-time calculation method of the distributed photovoltaic power generation meteorological elements includes the following steps:

[0056] S1. Select nearby sites based on power plant operation data and use the data from the nearby sites to repair the target site data;

[0057] The specific method of step S1 is:

[0058] S1-1. Integrate power data from photovoltaic power station operation data.

[0059] S1-2. Fill in missing data and replace outliers. Select the target point, calculate the distance between all power stations and the target power station, and select the 5 nearest neighboring power stations:

[0060]

[0061] Where r is the radius of the Earth (6378 km), Δlon is the longitude difference between the neighboring site and the target site, Δlat is the latitude difference between the neighboring site and the target site, lat1 is the latitude of the neighboring DPV site, and lat2 is the latitude of the target DPV site;

[0062] S1-3, based on the geographical distance between the adjacent reference site and the target site, calculate the weight w(x0, x i ), the weight is a function of the inverse of the feature distance, and the calculation formula is:

[0063]

[0064] in, f iis the feature vector of the neighboring site, f is the feature vector of the target site, d(x0,x i ) is the characteristic distance between the neighboring site and the target site;

[0065] S1-4, using the measured value y(x i ) Calculate the fitting value of the target power station That is the completed distributed photovoltaic power station power, the calculation formula is:

[0066]

[0067] S2, using the power data processed in S1, build a meteorological element calculation model, such as Figure 2 As shown, a model is established between numerical weather forecast, distributed photovoltaic site power and irradiance, and ambient temperature;

[0068] In step S2, the modeling process for irradiance data specifically includes the following sub-steps:

[0069] S2-1. Use the distributed photovoltaic power station power completed in S2 and the irradiance data in the numerical weather forecast as input data.

[0070] S2-2, calculate the forget gate output f based on the current input and the hidden state of the previous time step t , determines which information in the current memory unit needs to be forgotten, and its calculation formula is:

[0071] f t =σ(W f [h t-1 ,x t ]+b f );

[0072] Where: σ represents the sigmoid function, W f Represents the forget gate weight, h t-1 represents the hidden state from the previous time step t-1, x t represents the input of the current time step t, b f represents the forget gate bias;

[0073] S2-3, calculate the output value i of the input gate t and candidate memory unit values And add the results of the forget gate and the input gate to update the memory cell state c of the current time step t :

[0074] i t =σ(W i [h t-1 ,x t ]+b i );

[0075]

[0076] Where: W i Represents the input gate weight, W c represents the candidate unit state weight, c t-1 represents the cell state at the previous time step, b i represents the input gate bias, b c Represents the candidate unit state bias;

[0077] S2-4, according to the current input x t and the memory cell state c t , calculate the output value o of the output gate t , get the new hidden state h t , this state will be passed to the next time step as output and also serve as the output of the current time step:

[0078] o t =σ(W o [h t-1, x t ]+b o );

[0079] h t =o t *tanh(c t );

[0080] Where: W o represents the output gate weight, b o represents the output gate bias;

[0081] S2-5. Compare the output irradiance value with the actual irradiance value, and calculate whether the deviation d between the two meets the requirements. If d meets the requirements, obtain the specified weights and biases. If d does not meet the requirements, update the weights and biases through the back propagation algorithm, and repeat steps S2-2 to S2-4 until the end condition is met.

[0082] S2-6. Obtain calculated irradiation data.

[0083] In step S2, the modeling process for temperature data specifically includes the following sub-steps:

[0084] S3-1. Use the distributed photovoltaic power station power completed in S2 and the temperature data in the numerical weather forecast as input data.

[0085] S3-2. Repeat steps S2-2 to S2-4 until the end condition is met.

[0086] S3-3. Obtain calculated temperature data.

[0087] S3. Based on the model established in S2, the solar irradiance and temperature at the location of the distributed photovoltaic site are calculated using the real-time power and the current value of the numerical weather forecast as input.

[0088] Example 1

[0089] The 2022 historical power data of 96 distributed photovoltaic sites in a certain area, the 2022 historical power data of centralized photovoltaic power stations and the numerical weather forecasts provided by ECMWF and NASA were selected to verify the accuracy of the proposed method under different inputs, and the models containing power data and the models not containing power data were compared. Table 1 shows the performance indicators of the calculation results obtained by the method of the present invention. It can be observed that the calculation results of meteorological elements obtained by the method of the present invention are of good accuracy, and the Pearson correlation coefficient, MAE, MSE, and RMSE indicators are greatly improved. The above results are of great significance for the accurate measurement of meteorological information of distributed photovoltaic power generation clusters at low cost, and for the realization of intelligent operation and safe access of distributed photovoltaic power generation. Figure 3 and Figure 4 As shown, the irradiance and temperature calculated by the calculation method of the present invention are compared with the measured values, and the calculated data are slightly different from the measured values.

[0090] Table 1 Performance comparison of prediction results obtained by different prediction methods

[0091]

[0092] In summary, the method takes into account the comparison of different inputs when establishing the model; at the same time, in order to reduce the error of the calculation results, an adaptive weighted fusion algorithm is used to calculate the missing power data of distributed photovoltaics, and the meteorological parameters are calculated based on the neural network model, which is of great significance for realizing the intelligent operation and safe access of distributed photovoltaic power generation.

[0093] This embodiment provides a real-time calculation system for meteorological elements of distributed photovoltaic power generation, including:

[0094] A data repair module is used to select adjacent sites based on power plant operation data and use the data of adjacent sites to repair the target site data;

[0095] The model building module is used to build a meteorological element calculation model using the repaired power data, and establish a model between numerical weather forecast, distributed photovoltaic site power and irradiance, and ambient temperature;

[0096] The calculation module is used to calculate the solar irradiance and temperature at the location of the distributed photovoltaic site based on the constructed model, taking the real-time power and the current value of the numerical weather forecast as input.

[0097] Specifically, the data repair module is used to:

[0098] Integrate power data from photovoltaic power station operation data;

[0099] Fill in missing data and replace outliers. Select the target point and calculate the distances of all power stations from the target power station. Select the five nearest neighboring power stations.

[0100] Calculate the weight of each reference power station, the weight is a function of the inverse of the characteristic distance;

[0101] The measured values of the reference power station are used to calculate the fitting values of the target power station to complete the power of the distributed photovoltaic power station.

[0102] Optionally, the model building module adopts a neural network model, including a forget gate, an input gate, and an output gate, for dynamically updating the memory unit state and the hidden state.

[0103] Optionally, the calculation module is used to optimize the model parameters through a back propagation algorithm according to the deviation between the model output result and the actual value until the accuracy requirements are met.

[0104] The basic principles, main features, and advantages of the present invention are shown and described above. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The above embodiments and descriptions are merely illustrative of the principles of the present invention. Various changes and modifications may be made to the present invention without departing from the spirit and scope of the present invention. Such changes and modifications are intended to fall within the scope of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A real-time calculation method for meteorological elements of distributed photovoltaic power generation, characterized in that: The following steps are involved: S1. Selecting neighboring sites based on the operating data of the target power station, and using the data of the neighboring sites to complete the missing power data of the target site and replace abnormal values; S2. Using the completed power data and irradiance data from the numerical weather forecast, a meteorological element calculation model is constructed to establish a mapping relationship between the numerical weather forecast, the power of distributed photovoltaic sites, solar irradiance, and ambient temperature; S3. Using the real-time power data and the current value of the numerical weather forecast as input, the solar irradiance and ambient temperature at the location of the target power station are calculated in real time using the meteorological element calculation model.

2. The real-time calculation method of meteorological elements of distributed photovoltaic power generation according to claim 1 is characterized in that: Step S1 specifically includes: S1-1. Integrate power data from the photovoltaic power station operation data in the region; S1-2. Fill in the missing data of the target power station and replace outliers. Select the target point x0, calculate the distance between all power stations and the target power station, and select the 5 closest power stations as adjacent reference points. The distance calculation formula is: Where r is the radius of the earth, Δlon is the longitude difference between the neighboring site and the target site, Δlat is the latitude difference between the neighboring site and the target site, lat1 is the latitude of the neighboring DPV site, and lat2 is the latitude of the target DPV site; S1-3, based on the geographical distance between the adjacent reference site and the target site, calculate the weight w(x0, x i ), the weight is a function of the inverse of the feature distance, and the calculation formula is: Among them, f i is the feature vector of the neighboring site, f is the feature vector of the target site, d(x0,x i ) is the characteristic distance between the neighboring site and the target site; S1-4, using the measured value y(x i ) Calculate the fitting value of the target power station That is the completed distributed photovoltaic power station power, the calculation formula is:

3. The real-time calculation method of meteorological elements for distributed photovoltaic power generation according to claim 1, characterized in that: In step S2, the modeling process for solar radiation data specifically includes the following sub-steps: S2-1. Use the completed distributed photovoltaic power station power and irradiance data from numerical weather forecast as input data; S2-2, calculate the forget gate output f based on the current input and the hidden state of the previous time step t , determines which information in the current memory unit needs to be forgotten, and its calculation formula is: f t =σ(W f [h t-1 ,x t ]+b f ); Where: σ represents the sigmoid function, W f Represents the forget gate weight, h t-1 represents the hidden state from the previous time step t-1, x t represents the input of the current time step t, b f represents the forget gate bias; S2-3. Calculate the output value i of the input gate t and candidate memory cell values And add the results of the forget gate and the input gate to update the memory cell state c of the current time step t : i t =σ(W i [h t-1 ,x t ]+b i ); Where: W i Represents the input gate weight, W c represents the candidate unit state weight, c t-1 represents the cell state at the previous time step, b i represents the input gate bias, b c Represents the candidate unit state bias; S2-4, according to the current input x t and the memory cell state c t , calculate the output value o of the output gate t , get the new hidden state h t , this state will be passed to the next time step as output and also serve as the output of the current time step: the t =σ(W o [h t-1 ,x t ]+b o ); h t =o t *tanh(c t ); Where: W o represents the output gate weight, b o represents the output gate bias; S2-5. Compare the output irradiance value with the actual irradiance value and calculate whether the deviation d between the two meets the requirements. If d meets the requirements, obtain the specified weights and biases. If d does not meet the requirements, update the weights and biases through the back propagation algorithm and repeat steps S2-2 to S2-4 until the end condition is met. S2-6. Obtain calculated irradiation data.

4. The real-time calculation method of meteorological elements for distributed photovoltaic power generation according to claim 3 is characterized in that: In step S2, the modeling process for temperature data specifically includes the following sub-steps: S3-1, using the completed distributed photovoltaic power station power and temperature data from numerical weather forecasting as input data; S3-2, repeat steps S2-2 to S2-4 until the end condition is met; S3-3. Obtain calculated temperature data.

5. The real-time calculation method of meteorological elements for distributed photovoltaic power generation according to claim 1, characterized in that: The meteorological element calculation model is a LSTM-based neural network model.

6. A real-time calculation system for meteorological elements of distributed photovoltaic power generation, characterized in that: include: A data repair module is used to select adjacent sites based on power plant operation data and use the data of adjacent sites to repair the target site data; The model building module is used to build a meteorological element calculation model using the repaired power data, and establish a model between numerical weather forecast, distributed photovoltaic site power and irradiance, and ambient temperature; The calculation module is used to calculate the solar irradiance and temperature at the location of the distributed photovoltaic site based on the constructed model, taking the real-time power and the current value of the numerical weather forecast as input.

7. An electronic device, characterized in that: include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the steps of the method for real-time calculation of meteorological elements of distributed photovoltaic power generation as described in any one of claims 1 to 5 are implemented.

8. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of the real-time calculation method of distributed photovoltaic power generation meteorological elements according to any one of claims 1 to 5.

Images