Energy storage optimization method and system for photovoltaic power generation equipment

Abstract

The invention discloses an energy storage optimization method and system for photovoltaic power generation equipment, and relates to the technical field of photovoltaic power generation energy storage, and the method comprises the steps: obtaining real-time meteorological data of a target region, and predicting the power generation power of the photovoltaic power generation equipment in a future preset time period of the target region according to the real-time meteorological data; predicting the electricity consumption power of the target area in a future preset time period according to the historical electricity consumption power of the target area, and judging whether the target area has residual electricity or not according to the generated power and the electricity consumption power in the preset time period; if yes, storage optimization is carried out on the residual power of the photovoltaic power generation equipment; thus, the short-term photovoltaic power generation model can be trained according to the actual situation, the photovoltaic power generation power in the short term can be predicted for energy storage, it is ensured that the prediction result is reasonable, meanwhile, energy storage and power utilization cannot be balanced, and the power loss is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of photovoltaic power generation energy storage, and particularly relates to an energy storage optimization method and system for photovoltaic power generation equipment. Background Art

[0002] The energy storage of photovoltaic power generation equipment refers to the process of storing the excess power generated by the photovoltaic power generation equipment; by storing the excess power when the photovoltaic power generation is strong and connecting the stored power to the power grid for use when the photovoltaic power generation is weak, the efficient utilization of photovoltaic power generation is realized;

[0003] In order to store the power of the photovoltaic power generation equipment in time and reduce the power loss, generally, the power generation power of the photovoltaic power generation equipment in the target area is predicted, and combined with the power consumption power of the target area, it is judged whether there is still surplus power after the power generation power of the photovoltaic power generation equipment meets the power consumption demand. If there is surplus, the excess power is stored in time;

[0004] However, the existing method for predicting the photovoltaic power generation in the target area is often based on a long-term power generation model for prediction. This prediction method leads to a large deviation in the prediction of the photovoltaic power generation power, resulting in unreasonable prediction results, and at the same time, the energy storage and power consumption cannot reach balance, and even increases the power loss. Summary of the Invention

[0005] The object of the present invention is to solve the above-mentioned problems of unreasonable prediction results, which also lead to the inability to balance energy storage and power consumption, and even increase power loss, and provide an energy storage optimization method and system for photovoltaic power generation equipment.

[0006] In the first aspect of the implementation of the present invention, an energy storage optimization method for photovoltaic power generation equipment is first proposed, and the method includes:

[0007] Obtain the real-time meteorological data of the target area, and predict the power generation power of the photovoltaic power generation equipment in the target area within a preset time period in the future according to the real-time meteorological data;

[0008] Predict the power consumption power of the target area within a preset time period in the future according to the historical power consumption power of the target area, and judge whether there will be surplus power in the target area according to the power generation power and the power consumption power within the preset time period;

[0009] If there is surplus, optimize the storage of the surplus power of the photovoltaic power generation equipment,

[0010] Optionally, the step of predicting the power generation power of the photovoltaic power generation equipment in the target area within a preset time period in the future according to the real-time meteorological data is:

[0011] Obtaining historical meteorological data of the target area and historical photovoltaic power generation corresponding to the photovoltaic power generation equipment, the historical meteorological data including solar radiation, cloud coverage, and wind speed; the historical photovoltaic power generation is consistent with the timestamp of the historical meteorological data;

[0012] Clustering the historical meteorological data, and determining that the historical meteorological data is divided into a plurality of clusters according to the clustering result, each cluster corresponding to a label and a meteorological mode; the meteorological mode includes cloudy, clear and overcast;

[0013] Obtaining a photovoltaic power generation power prediction model for different meteorological modes according to the meteorological mode of the historical meteorological data and the historical photovoltaic power generation power;

[0014] Acquire real-time meteorological data of the target area, cluster the real-time meteorological data to determine the meteorological mode to which it belongs, and input the real-time meteorological data into the photovoltaic power generation prediction model corresponding to the meteorological mode to predict the power generation of photovoltaic power generation equipment within a preset time period in the future.

[0015] Optionally, the step of clustering the historical meteorological data to divide the historical meteorological data into a plurality of clusters is:

[0016] Clustering the historical meteorological data by different clustering methods, each clustering method divides the historical meteorological data into a plurality of cluster clusters; the different clustering methods include K-means clustering, hierarchical clustering, and Gaussian mixture clustering;

[0017] The historical photovoltaic data that do not overlap in each cluster in all clustering methods are marked as the first photovoltaic data, and the first photovoltaic data are integrated through the Hungarian algorithm. The Euclidean distance from the first photovoltaic data to each cluster is calculated as the matching cost, and it is classified into the cluster corresponding to the minimum matching cost, and the historical photovoltaic data is divided into the corresponding cluster;

[0018] The steps of obtaining a photovoltaic power generation power prediction model in different meteorological modes according to the meteorological mode of the historical meteorological data and the historical photovoltaic power generation are as follows:

[0019] For the data sets corresponding to the clusters of each meteorological mode, the data sets are respectively input into the PSO-VMD model for training, and the predicted power generation of the photovoltaic power generation equipment is output, until the difference between the predicted power generation output by the PSO-VMD model and the historical photovoltaic power generation is less than the preset power generation threshold, and the photovoltaic power generation prediction model of the corresponding meteorological mode is obtained.

[0020] Optionally, the steps of predicting the power consumption of the target area in a future preset time period according to the historical power consumption of the target area and determining whether there will be remaining power in the target area according to the power generation power and the power consumption in the preset time period are as follows:

[0021] Obtain the power consumption of the target area at the same time within the preset time period every day in the recent preset cycle, obtain the daily power consumption in the recent preset cycle, and use the box plot method to eliminate the abnormally fluctuating power consumption to obtain the available power consumption;

[0022] Calculate the mean and standard deviation of the available power consumption, and add the standard deviation to the mean as the power consumption of the target area in the future preset time period;

[0023] Subtract the power consumption from the power generation power in the preset time period. If the result is not greater than 0, there will be no remaining power in the target area; then all the power generated by the photovoltaic power generation equipment is connected to the power grid of the target area; if the result is greater than 0, there will be remaining power in the target area, and the power generation of the photovoltaic power generation equipment will be stored and optimized.

[0024] Optionally, the steps of storing and optimizing the remaining power of the photovoltaic power generation equipment are as follows:

[0025] Obtain the total number of charge and discharge cycles of each energy storage device of the photovoltaic power generation equipment, and divide the total number of charge and discharge cycles by the preset number of charge and discharge cycles to obtain the charge and discharge life ratio;

[0026] Obtain the average storage response time when the energy storage device stores electricity historically and the average release response time when the electricity is released, and add the storage response time and the release response time as the total response time of the energy storage device; normalize the total response time of each energy storage device and map it to the 0-1 interval as the response time ratio of the energy storage device;

[0027] Obtain the current storable power capacity of each energy storage device, and normalize the storable power capacity to obtain the storage capacity ratio of the energy storage device;

[0028] Calculate the energy storage priority value of the energy storage device according to the charge and discharge life ratio, response time ratio and storage capacity ratio of the energy storage device. The calculation formula is: W = Sd - Sx - Sz, where W is the energy storage priority value, and Sd, Sx and Sz are the storage capacity ratio, charge and discharge life ratio and response time ratio of the energy storage device respectively;

[0029] Store the remaining power of the photovoltaic power generation equipment in the energy storage device corresponding to the maximum energy storage priority value to optimize the storage of the remaining power of the photovoltaic power generation equipment.

[0030] In the second aspect of the implementation of the present invention, an energy storage optimization system for photovoltaic power generation equipment is proposed. The system includes:

[0031] Power generation power module: Obtain real-time meteorological data of the target area, and predict the power generation power of the photovoltaic power generation equipment in the target area within a preset time period in the future according to the real-time meteorological data;

[0032] Remaining power judgment module: Predict the power consumption of the target area within a preset time period according to the historical power consumption of the target area, and judge whether there will be remaining power in the target area according to the power generation power and the power consumption within the preset time period;

[0033] Energy storage optimization module: If there is remaining power, store and optimize the remaining power of the photovoltaic power generation equipment.

[0034] Optionally, the power generation power module includes:

[0035] Meteorological data module: Obtain the historical meteorological data of the target area and the corresponding historical photovoltaic power generation power of the photovoltaic power generation equipment. The historical meteorological data includes solar radiation, cloud cover, and wind speed; the time stamps of the historical photovoltaic power generation power and the historical meteorological data are consistent;

[0036] Clustering analysis module: Cluster the historical meteorological data, and determine that the historical meteorological data is divided into several clustering clusters according to the clustering results. Each clustering cluster corresponds to a label and a meteorological pattern; the meteorological patterns include cloudy, sunny, and overcast;

[0037] Prediction model module: Obtain the photovoltaic power generation power prediction models for different meteorological patterns according to the meteorological patterns of the historical meteorological data and the historical photovoltaic power generation power;

[0038] Power generation power prediction module: Obtain the real-time meteorological data of the target area, judge the meteorological pattern to which the real-time meteorological data belongs by clustering, and input the real-time meteorological data into the photovoltaic power generation power prediction model belonging to the corresponding meteorological pattern to predict the power generation power of the photovoltaic power generation equipment within a preset time period in the future.

[0039] Optionally, the clustering analysis module includes:

[0040] Initial clustering module: Cluster the historical meteorological data by different clustering methods. Each clustering method divides the historical meteorological data into several clustering clusters; the different clustering methods include K-means clustering, hierarchical clustering, and Gaussian mixture clustering;

[0041] Integration module: Mark the historical photovoltaic data that does not overlap in each clustering cluster of all clustering methods as the first photovoltaic data. Integrate the first photovoltaic data through the Hungarian algorithm, calculate the Euclidean distance from the first photovoltaic data to each assigned clustering cluster as the matching cost, classify it into the clustering cluster corresponding to the minimum matching cost, and divide the historical photovoltaic data into the corresponding clustering clusters;

[0042] The application of the prediction model module is as follows: For the datasets corresponding to the clustering clusters of each meteorological pattern, input the datasets into the PSO-VMD model for training respectively, and output the predicted power generation of the photovoltaic power generation device until the difference between the predicted power generation output by the PSO-VMD model and the historical photovoltaic power generation is less than the preset power generation threshold, and obtain the photovoltaic power generation prediction model for the corresponding meteorological pattern.

[0043] Optionally, the remaining power judgment module includes:

[0044] Available power consumption module: Obtain the power consumption within the same time period as the preset time period every day in the target area in the most recent preset period, obtain the power consumption every day in the most recent preset period, and use the box plot method to eliminate the abnormally fluctuating power consumption to obtain the available power consumption;

[0045] Power consumption module: Calculate the mean and standard deviation of the available power consumption, and add the mean and the standard deviation as the power consumption of the target area in the future preset time period;

[0046] Power judgment module: Subtract the power consumption from the power generation within the preset time period. If the result is not greater than 0, there will be no remaining power in the target area; then all the power generated by the photovoltaic power generation device is connected to the power grid of the target area; if the result is greater than 0, there will be remaining power in the target area, and then optimize the storage of the power generated by the photovoltaic power generation device.

[0047] Optionally, the energy storage optimization module includes:.

[0048] Charge and discharge life ratio module: Obtain the total number of times each energy storage device of the photovoltaic power generation device has been charged and discharged currently, and divide the total number of times of charge and discharge by the preset number of charge and discharge times to obtain the charged and discharged life ratio;

[0049] Response time ratio module: Obtain the average storage response time when the energy storage device stores electricity historically and the average release response time when the electricity is released, and add the storage response time and the release response time as the total response time of the energy storage device; Normalize the total response time of each energy storage device and map it to the 0-1 interval as the response time ratio of the energy storage device;

[0050] Storage capacity ratio module: Obtain the current storable power capacity of each energy storage device, and normalize the storable power capacity to obtain the storage capacity ratio of the energy storage device;

[0051] Energy storage priority value module: Calculate the energy storage priority value of the energy storage device according to the charged and discharged life ratio, response time ratio, and storage capacity ratio of the energy storage device. The calculation formula is: W = Sd - Sx - Sz, where W is the energy storage priority value, and Sd, Sx, and Sz are the storage capacity ratio, charged and discharged life ratio, and response time ratio of the energy storage device respectively;

[0052] Energy storage determination module: Store the remaining power of the photovoltaic power generation device in the energy storage device corresponding to the maximum energy storage priority value, and optimize the storage of the remaining power of the photovoltaic power generation device..

[0053] Advantages of the present invention:

[0054] The present invention provides an energy storage optimization method and system for a photovoltaic power generation device. By obtaining real-time meteorological data of a target area and predicting the power generation power of the photovoltaic power generation device in the target area within a preset time period in the future according to the real-time meteorological data; predicting the power consumption power of the target area within the preset time period according to the historical power consumption power of the target area, and judging whether there will be remaining power in the target area according to the power generation power and the power consumption power within the preset time period; if there is remaining power, optimize the storage of the remaining power of the photovoltaic power generation device; in this way, it is possible to train a short-term photovoltaic power generation model according to the actual situation, predict the short-term photovoltaic power generation power for energy storage, ensure that the prediction result is reasonable, and at the same time, it is impossible to achieve a balance between energy storage and power consumption, reducing power loss. Description of the drawings

[0055] The present invention will be further described below with reference to the accompanying drawings.

[0056] Figure 1 It is a flowchart of an energy storage optimization method for a photovoltaic power generation device;

[0057] Figure 2 It is a framework diagram of an energy storage optimization system for a photovoltaic power generation device. Detailed implementation manners

[0058] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0059] Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the scope of protection of the present invention.

[0060] An embodiment of the present invention provides an energy storage optimization method for a photovoltaic power generation device. Refer to Figure 1 , Figure 1 which is a flowchart of an energy storage optimization method for a photovoltaic power generation device provided by an embodiment of the present invention. The method includes the following steps:

[0061] Obtain the real-time meteorological data of the target area, and predict the power generation power of the photovoltaic power generation device in the target area within a preset time period in the future according to the real-time meteorological data;

[0062] Predict the power consumption of the target area within a preset time period in the future according to the historical power consumption of the target area, and judge whether there will be surplus power in the target area according to the power generation power and the power consumption within the preset time period;

[0063] If there is surplus, optimize the storage of the surplus power of the photovoltaic power generation device.

[0064] Based on the energy storage optimization method for a photovoltaic power generation device provided by an embodiment of the present invention, through the above method, it is possible to train a short-term photovoltaic power generation model according to the actual situation, predict the photovoltaic power generation power in the short term for energy storage, ensure that the prediction result is reasonable, and at the same time make the energy storage and power consumption unable to reach balance, reducing power loss.

[0065] In one embodiment, the step of predicting the power generation power of the photovoltaic power generation device in the target area within a preset time period in the future according to the real-time meteorological data is as follows:

[0066] Obtain the historical meteorological data of the target area and the corresponding historical photovoltaic power generation power of the photovoltaic power generation device. The historical meteorological data includes solar radiation, cloud cover, and wind speed; the time stamps of the historical photovoltaic power generation power and the historical meteorological data are consistent;

[0067] Cluster the historical meteorological data, and determine that the historical meteorological data is divided into several clustering clusters according to the clustering result. Each clustering cluster corresponds to a label and a meteorological pattern; the meteorological patterns include cloudy, sunny, and overcast;

[0068] Obtain the photovoltaic power generation power prediction models for different meteorological patterns according to the meteorological patterns of the historical meteorological data and the historical photovoltaic power generation power;

[0069] Obtain the real-time meteorological data of the target area, cluster the real-time meteorological data to determine the meteorological pattern it belongs to, and input the real-time meteorological data into the photovoltaic power prediction model belonging to the corresponding meteorological pattern to predict the power generation power of the photovoltaic power generation equipment within a preset future time period.

[0070] In one implementation, solar radiation, cloud cover, and wind speed are important meteorological factors affecting the power generation power of photovoltaic power generation equipment. Solar radiation directly affects the power generation efficiency of the photovoltaic panel. The stronger the radiation, the more electric energy the photovoltaic panel can convert. Cloud cover is a key factor affecting the radiation intensity. The thicker the cloud layer, the greater the reduction in the intensity of solar radiation, resulting in a decrease in photovoltaic power generation. Although wind speed has a relatively small direct impact on power generation efficiency, it affects the heat dissipation of the photovoltaic panel. To a certain extent, wind speed indirectly affects the power generation power by influencing the heat dissipation effect of the photovoltaic panel. A higher wind speed can help the panel dissipate heat, slow down the overheating phenomenon, and maintain a higher power generation efficiency. At the same time, when the wind speed is high, the dust on the surface of the solar panel is blown away, increasing the contact area with sunlight, thereby increasing power generation. The combination of these meteorological factors forms different meteorological patterns, such as sunny, cloudy, and overcast. Each pattern has different laws for the impact on photovoltaic power generation. In sunny weather, the cloud cover is low, the solar radiation intensity is high, and the wind speed is high, so the power generation power of photovoltaic power generation is usually high. In cloudy weather, the occlusion of the cloud layer will partially weaken the solar radiation, making the photovoltaic power generation power medium. In overcast or thick cloud conditions, the solar radiation is significantly weakened, and the improvement effect of wind speed on power generation may also be limited. Therefore, the photovoltaic power generation power is usually low. Therefore, based on these meteorological factors, photovoltaic power prediction models under different meteorological patterns can be established to help accurately predict the photovoltaic power generation capacity of the target area in the future for power management and optimization.

[0071] In one embodiment, the steps of clustering the historical meteorological data and dividing the historical meteorological data into several clustering clusters are as follows:

[0072] Cluster the historical meteorological data through different clustering methods. Each clustering method divides the historical meteorological data into several clustering clusters. The different clustering methods include K-means clustering, hierarchical clustering, and Gaussian mixture clustering.

[0073] Mark the historical photovoltaic data that does not overlap in each clustering cluster among all clustering methods as the first photovoltaic data. Integrate the first photovoltaic data through the Hungarian algorithm, calculate the Euclidean distance from the first photovoltaic data to each clustering cluster it is assigned to as the matching cost, classify it into the clustering cluster corresponding to the minimum matching cost, and divide the historical photovoltaic data into the corresponding clustering clusters.

[0074] The steps for obtaining the photovoltaic power prediction model for different meteorological patterns based on the meteorological patterns of the historical meteorological data and the historical photovoltaic power are as follows:

[0075] For the datasets corresponding to the clustering clusters of each meteorological pattern, the datasets are respectively input into the PSO-VMD model for training, and the predicted power generation of the photovoltaic power generation equipment is output until the difference between the predicted power generation output by the PSO-VMD model and the historical photovoltaic power is less than the preset power generation threshold, thereby obtaining the photovoltaic power prediction model for the corresponding meteorological pattern.

[0076] In one implementation, different clustering methods (such as K-means, hierarchical clustering, and Gaussian mixture clustering) are selected to cluster the historical meteorological data. Each clustering method has a different clustering logic. K-means assigns data to different clusters by minimizing the variance within the clusters; hierarchical clustering constructs a tree structure and gradually merges or splits the data according to the similarity; Gaussian mixture clustering estimates the probability distribution of the data by maximizing the likelihood function and divides it into clusters with multiple Gaussian distributions; each clustering method divides the historical meteorological data into several non-overlapping three clustering clusters, and the three clustering clusters respectively represent three different meteorological patterns: sunny, cloudy, and overcast.

[0077] In one implementation, the historical photovoltaic data that does not overlap in each clustering cluster among all clustering methods is marked as the first photovoltaic data. The first photovoltaic data is integrated through the Hungarian algorithm, and the Euclidean distance from the first photovoltaic data to each of the assigned clustering clusters is calculated as the matching cost, and it is classified into the clustering cluster corresponding to the minimum matching cost, thereby dividing the historical photovoltaic data into the corresponding clustering clusters.

[0078] For example, assume there is a set of historical meteorological data, T1, T2, T3, T4, T5, T6, T7, T8, T9, T10, T11, T12; each set of meteorological data includes solar radiation, cloud cover, and wind speed; assume that the K-means clustering method considers T1, T3, T5, T6 as one clustering cluster, and the meteorological pattern is sunny; T2, T7, T8, T9 as one clustering cluster, and the meteorological pattern is cloudy; T4, T10, T11, T12 as one clustering cluster, and the meteorological pattern is overcast; while hierarchical clustering considers T1 to belong to the cloudy weather cluster; the Gaussian mixture model considers T1 to belong to the sunny weather cluster; therefore, T1 is a controversial data, that is, the first photovoltaic data. The first photovoltaic data is integrated through the Hungarian algorithm, and the Euclidean distance from the first photovoltaic data to each of the assigned clustering clusters is calculated as the matching cost, and it is classified into the clustering cluster corresponding to the minimum matching cost, thereby dividing the historical photovoltaic data into the corresponding clustering clusters; the specific steps are as follows:

[0079] Step 1: Determine the cluster centers for each method: For each clustering method, calculate the center point (mean) of each cluster in the meteorological feature space; Example K-means cluster 1 sunny (center): sunny = (radiation mean, cloud cover mean, wind speed mean); Calculate the distance d from T1 to each cluster center: For each clustering method, calculate the Euclidean distance d from T1 to its assigned cluster and other clusters, and use the Euclidean distance as the matching cost;

[0080] Step 2: Construct the matching cost matrix:

[0081] Matrix structure: Rows: The assignment results of the disputed data point T1 in different clustering methods (K-means, hierarchical clustering, GMM (Gaussian mixture model));

[0082] Columns: Possible clustering clusters (sunny, cloudy, overcast);

[0083] Elements: The Euclidean distance from T1 to the corresponding cluster center, which is infinity (∞) if not assigned;

[0084] Example cost matrix is as follows:

[0085] Clustering method Sunny Cloudy Overcast K-means d1 ∞ ∞ Hierarchical clustering ∞ ∞ d2 GMM d3 ∞ ∞

[0086] Apply the Hungarian algorithm to solve the optimal matching:

[0087] Steps: Initialization: Construct the cost matrix (as shown in the above table);

[0088] Row reduction: Subtract the minimum value from each row to make at least one 0 in each row;

[0089] Column reduction: Subtract the minimum value from each column to make at least one 0 in each column;

[0090] Cover the zero elements: Cover all zero elements with the fewest lines;

[0091] Adjust the matrix: If the number of lines is less than the matrix order, adjust the uncovered elements and repeat until the optimal solution is found;

[0092] Assignment result: Select the positions of the zero elements to ensure that only one assignment is made in each row and each column;

[0093] Determine the final clustering label, select the cluster with the minimum total cost. For example, if d1 + d3 is less than d2, then T1 is assigned to the sunny cluster; if d1 + d3 is greater than d2, then T1 is assigned to the cloudy cluster.

[0094] It should be noted that the clustering labels are determined according to the actual situation; for example, if the labels are: clustering label = 1 (i.e., sunny cluster), clustering label = 2 (i.e., cloudy cluster), clustering label = 3 (i.e., overcast cluster), it is specifically determined according to the actual situation, without further limitation and elaboration.

[0095] Specific summary: Step connection and summary:

[0096] Data conflict identification: T1 is assigned to different clusters by different methods and needs to be integrated through the Hungarian algorithm;

[0097] Cost matrix construction: Quantify the allocation cost based on the Euclidean distance;

[0098] Hungarian algorithm optimization: Find the allocation scheme with the minimum total cost through row reduction, column reduction, and cover adjustment.

[0099] Label decision: Determine the final clustering label based on the principle of the minimum total cost.

[0100] Final result: T1 is integrated into the corresponding cluster and used as the input of the PSO-VMD model to improve the subsequent prediction accuracy.

[0101] In one implementation method, through the above method, the controversial photovoltaic data is assigned to reasonable clustering clusters, abnormal allocations are eliminated, the input data is optimized, and the robustness of the subsequent model is enhanced.

[0102] In one embodiment, the steps of obtaining the photovoltaic power prediction model for different weather patterns according to the weather pattern of the historical meteorological data and the historical photovoltaic power are as follows:

[0103] For the data sets corresponding to the clustering clusters of each weather pattern, the data sets are respectively input into the PSO-VMD model for training, and the predicted power generation of the photovoltaic power generation device is output until the difference between the predicted power generation output by the PSO-VMD model and the historical photovoltaic power is less than the preset power generation threshold, and the photovoltaic power prediction model for the corresponding weather pattern is obtained.

[0104] It should be noted that specifically, the PSO-VMD model is trained by clustering; the specific steps are, for example: Clustering training: According to the clustering labels, the data is divided into subsets clustering label = 1 (i.e., sunny cluster), clustering label = 2 (i.e., cloudy cluster), clustering label = 3 (i.e., overcast cluster), and the PSO-VMD model is independently trained;

[0105] First step: PSO optimizes VMD parameters (clustering training);

[0106] Objective: Dynamically optimize the number of modes K and the penalty factor α of VMD for different weather patterns.

[0107] Steps: For example, training the model for the sunny cluster:

[0108] Input data: Photovoltaic power time series data of T1, T3, T5, T6, T11.

[0109] PSO optimization parameters: Particle swarm size: 20 particles;

[0110] PSO initialization: Set the parameter range of the particle swarm (K ∈ [4, 6], α ∈ [1500, 2500]);

[0111] Fitness calculation: Use the reconstruction error of VMD decomposition as the optimization objective;

[0112] Parameter update: Iteratively optimize through the velocity and position formulas, and output the optimal parameters for the corresponding cluster (sunny cluster) (e.g., for cluster A, K = 5, α = 1900);

[0113] In one implementation, clustering training enables PSO to optimize parameters for different weather patterns (such as cloudy and sunny), improving the decomposition accuracy.

[0114] It should be noted that the clustering results directly affect the search space and fitness calculation of PSO, ensuring that the parameters match the weather patterns.

[0115] Second step: VMD signal decomposition and feature construction;

[0116] Objective: Decompose the corresponding photovoltaic power into time series signals, such as decomposing the photovoltaic power of the power time series data (such as T1, T3, T5, T6, T11) in the sunny cluster into time series signals; Extract multi-scale features;

[0117] The specific steps are, for example:

[0118] VMD decomposition: Use the parameters optimized by PSO to decompose the photovoltaic power generation signal into K IMFs.

[0119] Feature extraction: Calculate the statistics of the IMFs (such as mean, variance, energy);

[0120] Input vector: Integrate meteorological parameters, IMF features, and clustering labels (such as IMF1 mean, IMF2 variance, radiation, wind speed, cloud thickness, clustering label);

[0121] For example: Meteorological parameters: Normalized solar radiation (0.85), wind speed (0.12), cloud thickness (0.18); IMF statistics: IMF1 mean = 500, IMF2 variance = 25; Clustering label: The sunny cluster is encoded as 1; Then Input (input vector) = [IMF1 mean = 500, IMF2 variance = 25, radiation = 0.85, wind speed = 0.12, cloud thickness = 0.18, clustering label = 1];

[0122] Standardization processing: Perform Z-score standardization on continuous features (such as IMF mean, radiation);

[0123] Dataset division and standardization: Training set and test set: Divide by time sequence (the first 70% for training and the last 30% for testing);

[0124] Model training: Use the XGBoost regressor, input the feature vector, and output the power prediction for a future period of time;

[0125] Compare the output of the power prediction for a future period of time with the actual historical power generation. If the difference between the predicted power generation output by the PSO-VMD model and the historical photovoltaic power generation is less than the preset power generation threshold, a photovoltaic power generation prediction model for the corresponding meteorological pattern is obtained; if it is less, readjust the PSO parameter range or optimize the VMD decomposition strategy until the difference between the predicted power generation output by the PSO-VMD model and the historical photovoltaic power generation is less than the preset power generation threshold.

[0126] In one implementation method, through the above method, corresponding PSO-VMD photovoltaic power generation prediction models can be trained for different meteorological patterns, so as to predict the short-term photovoltaic power generation according to the meteorological pattern of the actual target area, making the prediction result more accurate and enabling energy storage in a more timely manner.

[0127] In one embodiment, the steps of obtaining the real-time meteorological data of the target area and predicting the power generation of the photovoltaic power generation equipment in the target area for a future preset time period according to the real-time meteorological data are as follows:

[0128] Obtain the real-time meteorological data of the target area for a current period of time, determine the meteorological pattern of the target area through clustering and the Hungarian algorithm, input the real-time meteorological data into the PSO-VMD photovoltaic power generation prediction model corresponding to the meteorological pattern, and predict the power generation for the future time period;

[0129] For example, assume that the real-time meteorological data of the target area reflects that the target area is currently in a sunny meteorological pattern; input the real-time meteorological data into the PSO-VMD photovoltaic power generation prediction model corresponding to the sunny meteorological pattern, and predict the power generation for the future time period; to achieve the prediction of the power generation of the photovoltaic power generation equipment in the target area for a future preset time period.

[0130] In one embodiment, the steps of predicting the power consumption of the target area for a future preset time period according to the historical power consumption of the target area and determining whether there will be surplus power in the target area according to the power generation and the power consumption during the preset time period are as follows:

[0131] Obtain the power consumption of the target area at the same time within the preset time period every day in the recent preset cycle, obtain the daily power consumption in the recent preset cycle, and use the box plot method to eliminate the abnormally fluctuating power consumption to obtain the available power consumption;

[0132] Calculate the mean and standard deviation of the available power consumption, and add the standard deviation to the mean as the power consumption of the target area in the future preset time period;

[0133] Subtract the power consumption from the power generation in the preset time period. If the result is not greater than 0, there will be no remaining power in the target area; then all the power generated by the photovoltaic power generation equipment will be connected to the power grid of the target area; if the result is greater than 0, there will be remaining power in the target area, and the power generation of the photovoltaic power generation equipment will be stored and optimized.

[0134] It should be noted that the preset cycle and preset time period are set by professionals according to the actual situation. For example, assume that it is necessary to predict the photovoltaic power generation power at 9-10 o'clock in the target area; obtain the power consumption at 9-10 o'clock every day in the past month in the target area to obtain the historical power consumption set; and use the box plot method (Boxplot) to identify and eliminate the abnormal fluctuations in the historical power consumption set. The box plot method can set the upper and lower boundaries by calculating the quartiles (Q1 and Q3) (usually Q1 - 1.5IQR and Q3 + 1.5IQR, where IQR is the interquartile range). Any data points outside this range can be regarded as abnormal fluctuations and eliminated; for example: in the past month, collect the power consumption at 9-10 o'clock every day (for example: 500, 520, 510, 530, 490, 505,...); identify and eliminate the outliers (such as eliminating the two data points 500 and 550 that deviate from the mean) through the box plot method; after eliminating the abnormal fluctuations, a set of reliable historical power consumption data is obtained, and these data are the "available power consumption".

[0135] In one implementation method, the advantage of using the box plot method (Boxplot) to identify and eliminate the abnormal fluctuations in the historical power consumption data is that it can intuitively display the data distribution, including the upper and lower quartiles and outliers (outliers), so as to help quickly identify the fluctuations that do not conform to the conventional pattern. By eliminating these abnormal data, the interference of data noise on the prediction model can be reduced, and the accuracy and stability of the prediction can be improved. In addition, the box plot method does not depend on the data distribution assumption, has strong adaptability, is applicable to various different types of data sets, especially in the case of irregular fluctuations and extreme values, and can effectively improve the data quality.

[0136] In one implementation, by calculating the mean and standard deviation of the available power consumption and adding the mean to the standard deviation to determine the power consumption of the target area within a preset future time period, the volatility and change trend of historical data can be effectively captured. The mean provides the central position of the power consumption, while the standard deviation reflects the fluctuation range of the power consumption. Adding the mean and the standard deviation can better estimate the range of power consumption that may occur in the future. Especially considering a certain degree of volatility, this method makes the prediction results more realistic and robust. In this way, overly idealized predictions can be effectively avoided, and at the same time, considering the natural fluctuations of the power consumption, it helps to optimize the power dispatching and energy storage strategies to ensure that the prediction results are closer to the actual situation.

[0137] It should be noted that by subtracting the power consumption of the target area from the power generation within the preset time period, it can be determined whether the power is sufficient. If the difference is not greater than 0, it means that the power consumption demand of the target area has been fully covered, and all the power generated by the photovoltaic power generation equipment will be connected to the power grid to meet the regional power consumption demand; on the contrary, if the difference is greater than 0, it means that the power demand of the target area has not been fully met, and the photovoltaic power generation equipment will generate surplus power. At this time, in order to avoid waste, through the storage optimization strategy, the excess power can be stored for use when the future power consumption demand is high. This not only effectively utilizes the surplus power of photovoltaic power generation, but also improves the flexibility of power dispatching and ensures the stability of energy supply.

[0138] In one embodiment, the steps for optimizing the storage of the surplus power of the photovoltaic power generation equipment are as follows:

[0139] Obtain the total number of charge and discharge cycles that each energy storage device of the photovoltaic power generation equipment has currently undergone, and divide the total number of charge and discharge cycles by the preset total number of charge and discharge cycles to obtain the charge and discharge life ratio;

[0140] Obtain the average storage response time when the energy storage device stores power historically and the average release response time when the power is released, and add the storage response time and the release response time as the total response time of the energy storage device; normalize the total response times of each energy storage device and map them to the 0-1 interval as the response time ratio of the energy storage device;

[0141] Obtain the current storable power capacity of each energy storage device, and normalize the storable power capacity to obtain the storage capacity ratio of the energy storage device;

[0142] Calculate the energy storage priority value of the energy storage device according to the charge and discharge life ratio, response time ratio, and storage capacity ratio of the energy storage device. The calculation formula is: W = Sd - Sx - Sz, where W is the energy storage priority value, and Sd, Sx, and Sz are the storage capacity ratio, charge and discharge life ratio, and response time ratio of the energy storage device respectively;

[0143] The surplus power of the photovoltaic power generation equipment is stored in the energy storage device corresponding to the maximum energy storage priority value, and the storage of the surplus power of the photovoltaic power generation equipment is optimized.

[0144] It should be noted that the number of charge and discharge times of the energy storage device can be monitored and recorded in real time through the management system of the energy storage device, and each charge and discharge operation will be automatically recorded and counted by the system. Secondly, the storage response time of historical stored power and the response time when power is released can be obtained through the control system of the energy storage device. Usually the system will record the timestamp of each power storage and release, and calculate the average response time. The storage capacity of the energy storage device can be obtained through the technical parameters of the device or the real-time monitoring system, which usually reflects the current charging status of the device. The health status and remaining service life of the device can be evaluated through the built-in sensors or maintenance records of the device, and calculated in combination with the accumulated number of charge and discharge times and the aging of the device. All of these data can be collected and processed through the management platform, monitoring system and historical maintenance records of the device to ensure the accuracy and timeliness of the calculation process.

[0145] In one implementation, storing electricity in photovoltaic power generation equipment according to the above method can significantly improve the overall efficiency and life of the energy storage system. By comprehensively considering the charge-discharge life ratio, response time ratio and storage capacity ratio of the energy storage equipment, the system can accurately distribute the remaining electricity to the most suitable energy storage equipment, thereby maximizing the utilization rate of each energy storage equipment. First, this method can ensure that electricity is stored in devices with fast response time and large storage capacity, improve the efficiency and response speed of electricity storage, especially when the electricity demand peaks, the stored electricity can be released in time. Secondly, considering the charge-discharge life ratio helps to extend the service life of the energy storage equipment and avoid equipment damage or performance degradation caused by frequent charging and discharging. In addition, energy storage optimization can effectively reduce electricity waste, improve the overall economic benefits and sustainability of photovoltaic power generation systems, and provide strong support for the intelligent management and optimized scheduling of future energy systems.

[0146] Based on the same inventive concept, the embodiment of the present invention also provides an energy storage optimization system for photovoltaic power generation equipment. Figure 2 , Figure 2 A framework diagram of an energy storage optimization system for photovoltaic power generation equipment provided by an embodiment of the present invention, the system comprising:

[0147] Power generation module: obtains real-time meteorological data of the target area, and predicts the power generation of photovoltaic power generation equipment in the target area within a preset time period in the future based on the real-time meteorological data;

[0148] Surplus power judgment module: Predict the power consumption of the target area within a preset future time period based on the historical power consumption of the target area, and determine whether there will be surplus power in the target area according to the power generation and power consumption within the preset time period;

[0149] Energy storage optimization module: If there is a surplus, store and optimize the surplus power of the photovoltaic power generation equipment.

[0150] Based on the energy storage optimization system for photovoltaic power generation equipment provided by the embodiments of the present invention, through the above method, it is possible to train a short-term photovoltaic power generation model according to the actual situation, predict the photovoltaic power generation in the short term for energy storage, ensure that the prediction result is reasonable, and at the same time make the energy storage and power consumption unable to reach balance, reducing power loss.

[0151] In one embodiment, the power generation module includes:

[0152] Meteorological data module: Obtain the historical meteorological data of the target area and the corresponding historical photovoltaic power generation of the photovoltaic power generation equipment. The historical meteorological data includes solar radiation, cloud cover, and wind speed; the time stamps of the historical photovoltaic power generation and the historical meteorological data are consistent;

[0153] Clustering analysis module: Cluster the historical meteorological data, and determine that the historical meteorological data is divided into several clustering clusters according to the clustering results. Each clustering cluster corresponds to a label and a meteorological pattern; the meteorological patterns include cloudy, sunny, and overcast;

[0154] Prediction model module: Obtain the photovoltaic power generation prediction models for different meteorological patterns according to the meteorological patterns of the historical meteorological data and the historical photovoltaic power generation;

[0155] Power generation prediction module: Obtain the real-time meteorological data of the target area, cluster and judge the meteorological pattern to which the real-time meteorological data belongs, and input the real-time meteorological data into the photovoltaic power generation prediction model belonging to the corresponding meteorological pattern to predict the power generation of the photovoltaic power generation equipment within a preset future time period.

[0156] In one embodiment, the clustering analysis module includes:

[0157] Initial clustering module: Cluster the historical meteorological data by different clustering methods. Each clustering method divides the historical meteorological data into several clustering clusters; the different clustering methods include K-means clustering, hierarchical clustering, and Gaussian mixture clustering;

[0158] Integration module: Mark all non-overlapping historical photovoltaic data in each clustering method as the first photovoltaic data, integrate the first photovoltaic data through the Hungarian algorithm, calculate the Euclidean distance from the first photovoltaic data to each clustering cluster as the matching cost, classify it into the cluster corresponding to the minimum matching cost, and divide the historical photovoltaic data into the corresponding clustering cluster;

[0159] The application of the prediction model module is as follows: for the data sets corresponding to the clustering clusters of each meteorological mode, the data sets are respectively input into the PSO-VMD model for training, and the predicted power generation power of the photovoltaic power generation equipment is output, until the difference between the predicted power generation power output by the PSO-VMD model and the historical photovoltaic power generation power is less than the preset power generation power threshold, and the photovoltaic power generation power prediction model corresponding to the meteorological mode is obtained.

[0160] In one embodiment, the remaining power determination module includes:

[0161] Available power module: obtain the power consumption of the target area every day in the latest preset period that is consistent with the preset time period, obtain the power consumption of each day in the latest preset period, and use the box plot method to eliminate the abnormally fluctuating power consumption to obtain the available power consumption;

[0162] Power consumption module: calculates the mean and standard deviation of the available power consumption, and adds the mean to the standard deviation as the power consumption of the target area in a future preset time period;

[0163] Power judgment module: subtract the power consumption from the power generation in the preset time period. If the result is not greater than 0, there will be no surplus power in the target area; all the power generated by the photovoltaic power generation equipment will be connected to the power grid in the target area; if the result is greater than 0, there will be surplus power in the target area, and the power generation of the photovoltaic power generation equipment will be optimized for storage.

[0164] In one embodiment, the energy storage optimization module includes:

[0165] The charge-discharge life ratio module is used to obtain the total number of times each energy storage device of the photovoltaic power generation equipment has been charged and discharged, and divide the total number of times that have been charged and discharged by the total number of times that can be charged and discharged to obtain the charge-discharge life ratio;

[0166] Response time ratio module: obtains the average storage response time of the energy storage device when storing electricity and the average release response time when releasing electricity, and adds the storage response time and release response time to obtain the total response time of the energy storage device; normalizes the total response time of each energy storage device and maps it to the interval of 0-1 as the response time ratio of the energy storage device;

[0167] Storage capacity ratio module: Obtain the current storable power capacity of each energy storage device, and normalize the storable power capacity to obtain the storage capacity ratio of the energy storage device;

[0168] Energy storage priority value module: Calculate the energy storage priority value of the energy storage device according to the charged and discharged life ratio, response time ratio and storage capacity ratio of the energy storage device. The calculation formula is: W = Sd - Sx - Sz, where W is the energy storage priority value, and Sd, Sx and Sz are the storage capacity ratio, charged and discharged life ratio and response time ratio of the energy storage device respectively;

[0169] Energy storage determination module: Store the remaining power of the photovoltaic power generation device in the energy storage device corresponding to the maximum energy storage priority value, and optimize the storage of the remaining power of the photovoltaic power generation device..

[0170] The above has described an embodiment of the present invention in detail, but the content described is only a preferred embodiment of the present invention and cannot be considered as limiting the scope of implementation of the present invention. All equivalent changes and improvements made according to the scope of the present invention application should still fall within the scope covered by the patent of the present invention.

Claims

1. An energy storage optimization method for photovoltaic power generation equipment, characterized in that, It includes the following steps: Obtain the real-time meteorological data of the target area, and predict the power generation power of the photovoltaic power generation equipment in the target area within a preset future time period according to the real-time meteorological data; Predict the power consumption power of the target area within a preset future time period according to the historical power consumption power of the target area, and judge whether there will be surplus power in the target area according to the power generation power and the power consumption power within the preset time period; If there is a surplus, optimize the storage of the surplus power of the photovoltaic power generation equipment.

2. The energy storage optimization method for a photovoltaic power generation device according to claim 1, wherein The step of predicting the power generation power of the photovoltaic power generation equipment in the target area within a preset future time period according to the real-time meteorological data is as follows: Obtain the historical meteorological data of the target area and the corresponding historical photovoltaic power generation power of the photovoltaic power generation equipment. The historical meteorological data includes solar radiation, cloud cover, and wind speed; the time stamps of the historical photovoltaic power generation power and the historical meteorological data are consistent; Cluster the historical meteorological data, and determine that the historical meteorological data is divided into several clustering clusters according to the clustering result. Each clustering cluster corresponds to a label and a meteorological pattern; the meteorological patterns include cloudy, sunny, and overcast; Obtain the photovoltaic power generation power prediction models for different meteorological patterns according to the meteorological patterns of the historical meteorological data and the historical photovoltaic power generation power; Obtain the real-time meteorological data of the target area, cluster and judge the meteorological pattern to which the real-time meteorological data belongs, and input the real-time meteorological data into the photovoltaic power generation power prediction model belonging to the corresponding meteorological pattern to predict the power generation power of the photovoltaic power generation equipment within a preset future time period.

3. The energy storage optimization method for a photovoltaic power generation device according to claim 2, wherein The step of clustering the historical meteorological data and dividing the historical meteorological data into several clustering clusters is as follows: Cluster the historical meteorological data through different clustering methods. Each clustering method divides the historical meteorological data into several clustering clusters; the different clustering methods include K-means clustering, hierarchical clustering, and Gaussian mixture clustering; Mark the historical photovoltaic data that does not overlap in each clustering cluster among all the clustering methods as the first photovoltaic data, integrate the first photovoltaic data through the Hungarian algorithm, calculate the Euclidean distance from the first photovoltaic data to each clustering cluster it is assigned to as the matching cost, and classify it into the clustering cluster corresponding to the minimum matching cost, and divide the historical photovoltaic data into the corresponding clustering clusters; The step of obtaining the photovoltaic power generation power prediction models for different meteorological patterns according to the meteorological patterns of the historical meteorological data and the historical photovoltaic power generation power is as follows: For the data sets corresponding to the clustering clusters of each meteorological pattern, input the data sets into the PSO-VMD model for training respectively, and output the predicted power generation power of the photovoltaic power generation equipment until the difference between the predicted power generation power output by the PSO-VMD model and the historical photovoltaic power generation power is less than the preset power generation power threshold, and obtain the photovoltaic power generation power prediction model for the corresponding meteorological pattern.

4. A method for optimizing energy storage of a photovoltaic power generation device according to claim 1, characterized in that, The step of predicting the power consumption power of the target area within a preset future time period according to the historical power consumption power of the target area and judging whether there will be surplus power in the target area according to the power generation power and the power consumption power within the preset time period is as follows: Obtain the power consumption of the target area every day in the most recent preset period that is consistent with the preset time period, obtain the power consumption of each day in the most recent preset period, and use the box plot method to eliminate the power consumption with abnormal fluctuations to obtain the available power consumption; Calculate the mean and standard deviation of the available power, and add the mean plus the standard deviation as the power consumption of the target area in a future preset time period; The power generation power within the preset time period is subtracted from the power consumption power. If the result is not greater than 0, there will be no surplus power in the target area; all the power generated by the photovoltaic power generation equipment is connected to the power grid in the target area; if the result is greater than 0, there will be surplus power in the target area, and the power generation of the photovoltaic power generation equipment is optimized for storage.

5. A method for optimizing energy storage of a photovoltaic power generation device according to claim 1, characterized in that The steps to optimize the storage of surplus power from photovoltaic power generation equipment are: Obtain the total number of times each energy storage device of the photovoltaic power generation device has been charged and discharged, and divide the total number of times that have been charged and discharged by the preset total number of times that can be charged and discharged to obtain a charge and discharge life ratio; Obtain the average storage response time of the energy storage device when storing power in history and the average release response time when releasing power, and add the storage response time and the release response time as the total response time of the energy storage device; The total response time of each energy storage device is normalized and mapped to the range of 0-1 as the response time ratio of the energy storage device; Obtaining the current storable power capacity of each energy storage device, and normalizing the storable power capacity to obtain the storage capacity ratio of the energy storage device; The energy storage priority value of the energy storage device is calculated according to the charge and discharge life ratio, response time ratio and storage capacity ratio of the energy storage device. The calculation formula is: W = Sd-Sx-Sz, where W is the energy storage priority value, Sd, Sx and Sz are the storage capacity ratio, charge and discharge life ratio and response time ratio of the energy storage device respectively; The surplus power of the photovoltaic power generation equipment is stored in the energy storage device corresponding to the maximum energy storage priority value, and the storage of the surplus power of the photovoltaic power generation equipment is optimized.

6. An energy storage optimization system for a photovoltaic power generation device, characterized in that, The system includes a power generation module: acquiring real-time meteorological data of a target area, and predicting the power generation of photovoltaic power generation equipment in the target area within a preset time period in the future according to the real-time meteorological data; Remaining power judgment module: predicting the power consumption of the target area in a preset time period in the future according to the historical power consumption of the target area, and judging whether the target area will have surplus power according to the generated power and the power consumption in the preset time period; Energy storage optimization module: If there is surplus, the surplus power of the photovoltaic power generation equipment is stored and optimized.

7. The energy storage optimization system for a photovoltaic power generation device according to claim 6, characterized in that The power generation module comprises: Meteorological data module: obtains historical meteorological data of the target area and historical photovoltaic power generation corresponding to the photovoltaic power generation equipment, the historical meteorological data includes solar radiation, cloud coverage, and wind speed; the historical photovoltaic power generation is consistent with the timestamp of the historical meteorological data; Cluster analysis module: clustering the historical meteorological data, and determining that the historical meteorological data is divided into a number of clusters according to the clustering results, each cluster corresponding to a label and a meteorological mode; the meteorological mode includes cloudy, clear and overcast; Prediction model module: obtaining a photovoltaic power generation prediction model of different meteorological modes according to the meteorological mode of the historical meteorological data and the historical photovoltaic power generation; Power generation prediction module: obtains real-time meteorological data of the target area, clusters the real-time meteorological data to determine the meteorological mode to which it belongs, and inputs the real-time meteorological data into the photovoltaic power generation prediction model corresponding to the meteorological mode to predict the power generation of photovoltaic power generation equipment within a preset time period in the future.

8. The energy storage optimization system for a photovoltaic power generation device according to claim 7, wherein, The cluster analysis module comprises: Preliminary clustering module: clustering the historical meteorological data by different clustering methods, each clustering method divides the historical meteorological data into a number of clusters; the different clustering methods include K-means clustering, hierarchical clustering, and Gaussian mixture clustering; Integration module: Mark all non-overlapping historical photovoltaic data in each clustering method as the first photovoltaic data, integrate the first photovoltaic data through the Hungarian algorithm, calculate the Euclidean distance from the first photovoltaic data to each clustering cluster as the matching cost, classify it into the cluster corresponding to the minimum matching cost, and divide the historical photovoltaic data into the corresponding clustering cluster; The application of the prediction model module is as follows: for the data sets corresponding to the clustering clusters of each meteorological mode, the data sets are respectively input into the PSO-VMD model for training, and the predicted power generation power of the photovoltaic power generation equipment is output, until the difference between the predicted power generation power output by the PSO-VMD model and the historical photovoltaic power generation power is less than the preset power generation power threshold, and the photovoltaic power generation power prediction model corresponding to the meteorological mode is obtained.

9. The energy storage optimization system for a photovoltaic power generation device according to claim 6, characterized in that, The remaining power determination module comprises: Available power module: obtain the power consumption of the target area every day in the latest preset period that is consistent with the preset time period, obtain the power consumption of each day in the latest preset period, and use the box plot method to eliminate the abnormally fluctuating power consumption to obtain the available power consumption; Power consumption module: calculates the mean and standard deviation of the available power consumption, and adds the mean to the standard deviation as the power consumption of the target area in a future preset time period; Power judgment module: subtract the power consumption from the power generation in the preset time period. If the result is not greater than 0, there will be no surplus power in the target area; all the power generated by the photovoltaic power generation equipment will be connected to the power grid in the target area; if the result is greater than 0, there will be surplus power in the target area, and the power generation of the photovoltaic power generation equipment will be optimized for storage.

10. The energy storage optimization system for a photovoltaic power generation device according to claim 6, wherein, The energy storage optimization module includes: The charge-discharge life ratio module is used to obtain the total number of times each energy storage device of the photovoltaic power generation equipment has been charged and discharged, and divide the total number of times that have been charged and discharged by the total number of times that can be charged and discharged to obtain the charge-discharge life ratio; Response time ratio module: obtains the average storage response time of the energy storage device when storing electricity and the average release response time when releasing electricity, and adds the storage response time and release response time to obtain the total response time of the energy storage device; normalizes the total response time of each energy storage device and maps it to the interval of 0-1 as the response time ratio of the energy storage device; Storage capacity ratio module: Obtain the current storable power capacity of each energy storage device, and normalize the storable power capacity to obtain the storage capacity ratio of the energy storage device; Energy storage priority value module: Calculate the energy storage priority value of the energy storage device according to the charged and discharged life ratio, response time ratio and storage capacity ratio of the energy storage device. The calculation formula is: W = Sd - Sx - Sz, where W is the energy storage priority value, and Sd, Sx and Sz are the storage capacity ratio, charged and discharged life ratio, and response time ratio of the energy storage device respectively; Energy storage determination module: Store the remaining power of the photovoltaic power generation device in the energy storage device corresponding to the maximum energy storage priority value, and optimize the storage of the remaining power of the photovoltaic power generation device.

Images