Intelligent photovoltaic management method and application system based on BIM big data model

Abstract

The invention provides an intelligent photovoltaic management method and application system based on a BIM big data model, and the method comprises the steps: carrying out the data analysis of historical data of a target photovoltaic power station, obtaining a historical index value of the index data of the target photovoltaic power station under each weather type condition and historical photovoltaic power generation power corresponding to the historical index value; performing model training based on the historical index value of the index data under each weather type condition and the historical photovoltaic power generation power corresponding to the historical index value to obtain a target BIM big data photovoltaic prediction model; inputting the target index value of the index data of the target photovoltaic power station under the current weather type into the target BIM big data photovoltaic prediction model to obtain the current photovoltaic power generation power output by the target BIM big data photovoltaic prediction model; and carrying out equipment cleaning on operation equipment of the target photovoltaic power station based on the current photovoltaic power generation power. The power generation efficiency of the photovoltaic power station is improved.

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and in particular to an intelligent photovoltaic management method and application system based on a BIM big data model. Background Art

[0002] As the global demand for clean energy continues to rise, photovoltaic power generation has developed rapidly as a sustainable energy solution. The scale of photovoltaic power stations is expanding, from distributed small power stations to large centralized power stations, and they are playing an increasingly important role in the energy supply system. The photovoltaic equipment in these power stations, especially photovoltaic modules, current transformers and inverters, is the core part of converting solar energy into electrical energy. Its performance and efficiency directly affect the power generation and economic benefits of the entire power station.

[0003] At present, the management method for the operating equipment of photovoltaic power stations is mainly the sensor monitoring method, that is, installing corresponding power monitoring sensors in photovoltaic power stations, collecting the power generation of photovoltaic power stations through power monitoring sensors, and notifying manual cleaning of the operating equipment of photovoltaic power stations when the data collected by the power monitoring sensors is determined to be abnormal. In other words, the above method is to notify the management personnel to clean only after the power generation of photovoltaic power stations is abnormal, which makes the management of the operating equipment of photovoltaic power stations have obvious lags, resulting in power generation losses and potential damage to equipment, and reducing the power generation efficiency of photovoltaic power stations. Summary of the invention

[0004] The present invention provides an intelligent photovoltaic management method and application system based on the BIM big data model, which is used to solve the problem that the management of the operating equipment of the photovoltaic power station in the existing method has obvious lag, resulting in power generation losses and potential damage to the equipment, and to improve the power generation efficiency of the photovoltaic power station.

[0005] In a first aspect, the present invention provides an intelligent photovoltaic management method based on a BIM big data model, comprising:

[0006] Performing data analysis on the historical data of the target photovoltaic power station to obtain the historical index values ​​of the index data of the target photovoltaic power station under each weather type condition and the historical photovoltaic power generation power corresponding to the historical index values; the index data includes at least one of irradiance, temperature, humidity, wind speed and wind direction;

[0007] Model training is performed based on the historical indicator values ​​of the indicator data under each weather type condition and the historical photovoltaic power generation power corresponding to the historical indicator values, and the target BIM big data photovoltaic prediction model is obtained;

[0008] Inputting the target index value of the index data of the target photovoltaic power station under the current weather type into the target BIM big data photovoltaic prediction model to obtain the current photovoltaic power generation power output by the target BIM big data photovoltaic prediction model;

[0009] Based on the current photovoltaic power generation power, equipment cleaning is performed on operating equipment of the target photovoltaic power station; the operating equipment includes at least one of a photovoltaic component, a current transformer and an inverter.

[0010] According to the intelligent photovoltaic management method based on the BIM big data model provided by an embodiment of the present invention, the model training is performed based on the historical indicator values ​​of the indicator data under each weather type condition and the historical photovoltaic power generation power corresponding to the historical indicator values ​​to obtain the target BIM big data photovoltaic prediction model, including:

[0011] Taking each weather type, the historical indicator value of the indicator data under each weather type condition, and the historical photovoltaic power generation power corresponding to the historical indicator value as a sample, a data matrix is ​​constructed; wherein, in the data matrix, each weather type and the historical indicator value of the indicator data under each weather type condition are independent variables, and the historical photovoltaic power generation power corresponding to the historical indicator value is the dependent variable;

[0012] Taking each sample in the data matrix as a cluster, and constructing a cluster distance matrix based on the cluster distance between every two clusters in the data matrix;

[0013] Based on the cluster distance matrix, two clusters with the smallest cluster distance in the data matrix are merged until all clusters in the data matrix are merged to obtain a target cluster tree; in the target cluster tree, nodes represent clusters, edges between nodes represent cluster merging operations, the new cluster is the parent node of the two original clusters, and the two original clusters are respectively the two child nodes of the new cluster;

[0014] The target clustering tree is used as an initial BIM big data photovoltaic prediction model, and the initial BIM big data photovoltaic prediction model is optimized to obtain the target BIM big data photovoltaic prediction model.

[0015] According to the intelligent photovoltaic management method based on the BIM big data model provided by an embodiment of the present invention, the two clusters with the smallest cluster distance in the data matrix are merged based on the cluster distance matrix until all clusters in the data matrix are merged to obtain a target cluster tree, including:

[0016] Based on the cluster distance matrix, two clusters with the smallest cluster distance in the data matrix are merged until all clusters in the data matrix are merged to obtain an initial cluster tree;

[0017] Determining a cluster center of the initial cluster tree, and determining a cluster allocation distance between each cluster in the initial cluster tree and the cluster center;

[0018] The clusters with the same cluster allocation distance in the initial clustering tree are merged until all clusters in the initial clustering tree are merged to obtain the target clustering tree.

[0019] According to the intelligent photovoltaic management method based on the BIM big data model provided by an embodiment of the present invention, the initial BIM big data photovoltaic prediction model is optimized to obtain the target BIM big data photovoltaic prediction model, including:

[0020] Input the test data set into the initial BIM big data photovoltaic prediction model to obtain the predicted photovoltaic power generation of each test data in the test data set output by the initial BIM big data photovoltaic prediction model;

[0021] Based on the predicted photovoltaic power generation and the actual photovoltaic power generation of each test data, the initial BIM big data photovoltaic prediction model is optimized to obtain the target BIM big data photovoltaic prediction model.

[0022] According to the intelligent photovoltaic management method based on the BIM big data model provided by an embodiment of the present invention, the predicted photovoltaic power generation and the actual photovoltaic power generation based on each test data are used to optimize the initial BIM big data photovoltaic prediction model to obtain the target BIM big data photovoltaic prediction model, including:

[0023] Based on the predicted PV power and the actual PV power for each test data, the mean absolute error rate, root mean square, average variance, goodness of fit and daily accuracy of the power are determined;

[0024] Determine the confidence of the initial BIM big data photovoltaic prediction model based on the mean absolute error rate, root mean square, average variance, goodness of fit and daily accuracy of the generated power;

[0025] The cluster center of the initial BIM big data photovoltaic prediction model is adjusted based on the confidence level, and the confidence level of the initial BIM big data photovoltaic prediction model after the cluster center adjustment is cyclically obtained until the confidence level is greater than or equal to a preset threshold, so as to obtain the target BIM big data photovoltaic prediction model.

[0026] According to the intelligent photovoltaic management method based on the BIM big data model provided by an embodiment of the present invention, the cleaning of the operating equipment of the target photovoltaic power station based on the current photovoltaic power generation power includes:

[0027] For the case where the current photovoltaic power generation power is greater than the preset power generation power, and the power difference between the current photovoltaic power generation power and the preset power generation power is greater than the first preset difference:

[0028] If the power difference is within the first difference range, it is determined whether the operating temperature value of the photovoltaic component is within its preset operating temperature range; if not, a first cleaning instruction is sent to the cleaning device, so that the cleaning device cleans the photovoltaic component according to the first cleaning instruction; the cleaning device includes a vibration device, a mobile device and a water spraying device, and when cleaning, the vibration device, the mobile device and the water spraying device work together; if yes, it is determined whether the operating temperature value of the current transformer is within its preset operating temperature range; if not, a second cleaning instruction is sent to the cleaning device, so that the cleaning device cleans the current transformer according to the second cleaning instruction; if yes, a third cleaning instruction is sent to the cleaning device, so that the cleaning device cleans the inverter according to the third cleaning instruction;

[0029] If the power difference is within the second difference range, a fourth cleaning instruction is sent to the cleaning device, so that the cleaning device cleans the photovoltaic component, the current transformer and the inverter according to the fourth cleaning instruction;

[0030] According to the intelligent photovoltaic management method based on the BIM big data model provided by an embodiment of the present invention, the method further includes:

[0031] For the case where the current photovoltaic power generation power is greater than the preset power generation power, and the power difference between the current photovoltaic power generation power and the preset power generation power is greater than or equal to the first preset difference:

[0032] Sending a start instruction to a camera device corresponding to the target photovoltaic power station to obtain a first top image of the photovoltaic assembly, a second top image of the current transformer, and a third top image of the inverter acquired by the camera device based on the start instruction;

[0033] Performing image detection on the first top image, the second top image, and the third top image to obtain a target top image with an occluder;

[0034] A vibration instruction is sent to the cleaning device based on the target top image, so that the cleaning device removes obstructions on the top of the photovoltaic component, the current transformer or the inverter based on the vibration instruction.

[0035] In a second aspect, the present invention further provides an intelligent photovoltaic management system based on a BIM big data model, which is applied to the intelligent photovoltaic management method based on a BIM big data model as described in the first aspect; the intelligent photovoltaic management system based on a BIM big data model includes a data analysis module, a model training module, a power prediction module and an equipment management module;

[0036] A data analysis module is used to analyze the historical data of the target photovoltaic power station to obtain the historical index values ​​of the index data of the target photovoltaic power station under each weather type condition and the historical photovoltaic power generation power corresponding to the historical index values; the index data includes at least one of irradiance, temperature, humidity, wind speed and wind direction;

[0037] A model training module is used to perform model training based on the historical indicator values ​​of the indicator data under each weather type condition and the historical photovoltaic power generation power corresponding to the historical indicator values, so as to obtain the target BIM big data photovoltaic prediction model;

[0038] A power prediction module, used to input the target index value of the index data of the target photovoltaic power station under the current weather type into the target BIM big data photovoltaic prediction model, and obtain the current photovoltaic power generation power output by the target BIM big data photovoltaic prediction model;

[0039] The equipment management module is used to clean the operating equipment of the target photovoltaic power station based on the current photovoltaic power generation power; the operating equipment includes at least one of a photovoltaic module, a current transformer and an inverter.

[0040] In a third aspect, the present invention further provides an electronic device, comprising: a memory for storing a computer software program; and a processor for reading and executing the computer software program, thereby implementing any of the above-mentioned intelligent photovoltaic management methods based on the BIM big data model.

[0041] In a fourth aspect, the present invention further provides a non-transitory computer-readable storage medium, wherein a computer software program is stored in the storage medium. When the computer software program is executed by a processor, the intelligent photovoltaic management method based on the BIM big data model as described in any one of the above is implemented.

[0042] In a fifth aspect, the present invention also provides a computer program product, including a computer program, which, when executed by a processor, implements any of the above-mentioned intelligent photovoltaic management methods based on the BIM big data model.

[0043] The intelligent photovoltaic management method based on the BIM big data model provided by the embodiment of the present invention trains a target BIM big data photovoltaic prediction model jointly influenced by multiple indicator data through the indicator data under each weather type condition in the historical data and the corresponding historical photovoltaic power generation power. In the subsequent process, the indicator values ​​of the indicator data of the target photovoltaic power station under the current weather type are obtained in advance and input into the target BIM big data photovoltaic prediction model, so that the current photovoltaic power generation power of the photovoltaic power station can be obtained in advance, so that the operating equipment of the photovoltaic power station can be cleaned in advance according to the current photovoltaic power generation power, which solves the problem of management lag, avoids power generation losses and potential damage to equipment, and improves the power generation efficiency of the photovoltaic power station. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 It is a flow chart of an intelligent photovoltaic management method based on a BIM big data model provided by an embodiment of the present invention;

[0045] Figure 2 is a schematic diagram of the structure of an initial clustering tree provided by an embodiment of the present invention;

[0046] Figure 3 is a schematic diagram of the structure of a target clustering tree provided by an embodiment of the present invention;

[0047] Figure 4 It is a structural schematic diagram of an intelligent photovoltaic management system based on a BIM big data model provided by an embodiment of the present invention;

[0048] Figure 5 An embodiment diagram of an electronic device provided by an embodiment of the present invention;

[0049] Figure 6 An embodiment diagram of a computer-readable storage medium provided for an embodiment of the present invention. DETAILED DESCRIPTION

[0050] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. 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 those skilled in the art without creative work are within the scope of protection of the present invention.

[0051] In the description of the present invention, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Thus, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. In the description of the present invention, the meaning of "plurality" is two or more, unless otherwise clearly and specifically defined.

[0052] In the description of the present invention, the term "for example" is used to mean "used as an example, illustration or explanation". Any embodiment described as "for example" in the present invention is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is given to enable any technician in the field to implement and use the present invention. In the following description, details are listed for the purpose of explanation. It should be understood that a person of ordinary skill in the art can recognize that the present invention can be implemented without using these specific details. In other examples, well-known structures and processes will not be elaborated in detail to avoid obscuring the description of the present invention with unnecessary details. Therefore, the present invention is not intended to be limited to the embodiments shown, but is consistent with the widest scope consistent with the principles and features disclosed in the present invention.

[0053] See also Figure 1 , Figure 1 : is a flow chart of the intelligent photovoltaic management method based on the BIM big data model provided by the present invention. The executor of the intelligent photovoltaic management method based on the BIM big data model in the embodiment of the present invention is the photovoltaic management system. Therefore, the intelligent photovoltaic management method based on the BIM big data model includes:

[0054] Step 10, analyzing the historical data of the target photovoltaic power station to obtain the historical index values ​​of the index data of the target photovoltaic power station under each weather type condition and the historical photovoltaic power generation power corresponding to the historical index values.

[0055] Optionally, the target photovoltaic power station in the embodiment of the present invention may be any photovoltaic power station in a certain area, and the historical data may be the historical data of the target photovoltaic power station at a certain time point or within a certain period of time. Therefore, the photovoltaic management system performs data analysis on the historical data of the target photovoltaic power station to obtain the weather type in the historical data and the index data correlated with the photovoltaic power generation under the weather type, wherein the index data includes at least one of irradiance, temperature, humidity, wind speed and wind direction.

[0056] Therefore, the photovoltaic management system can obtain the historical indicator values ​​of the indicator data of the target photovoltaic power station under each weather type condition and the historical photovoltaic power generation power corresponding to the historical indicator values.

[0057] Step 20, based on the historical indicator values ​​of the indicator data under each weather type condition and the historical photovoltaic power generation power corresponding to the historical indicator values, a model training is performed to obtain a target BIM big data photovoltaic prediction model.

[0058] Furthermore, the photovoltaic management system continues to train the preset model with the historical indicator values ​​of the indicator data under each weather type condition and the historical photovoltaic power generation power corresponding to the historical indicator values ​​as training data to obtain the target BIM big data photovoltaic prediction model, wherein the preset model can be a pre-trained neural network model or a clustering model, such as a binary tree, a Huffman tree, etc.

[0059] The pre-trained model in the embodiment of the present invention is the cluster number, so it can be understood that the photovoltaic management system uses the historical indicator values ​​of the indicator data under each weather type condition and the historical photovoltaic power generation power corresponding to the historical indicator values ​​as training data to train the cluster number model to obtain the target BIM big data photovoltaic prediction model, as specifically described in steps 201 to 204.

[0060] Step 30, input the target indicator value of the indicator data of the target photovoltaic power station under the current weather type into the target BIM big data photovoltaic prediction model to obtain the current photovoltaic power generation power output by the target BIM big data photovoltaic prediction model.

[0061] Furthermore, the photovoltaic management system pre-acquires the target indicator value of the indicator data of the target photovoltaic power station under the current weather type. The target indicator value of the indicator data of the target photovoltaic power station under the current weather type can be obtained according to the weather forecast application or weather meteorological website at the current time.

[0062] Furthermore, the photovoltaic management system inputs the target indicator value of the indicator data of the target photovoltaic power station under the current weather type into the target BIM big data photovoltaic prediction model for power prediction, and obtains the current photovoltaic power generation power output by the target BIM big data photovoltaic prediction model.

[0063] Step 40: Clean the operating equipment of the target photovoltaic power station based on the current photovoltaic power generation power; the operating equipment includes at least one of a photovoltaic module, a current transformer and an inverter.

[0064] Furthermore, the photovoltaic management system compares the current photovoltaic power generation power with the preset power, and cleans the operating equipment of the target photovoltaic power station according to the comparison result, wherein the operating equipment includes photovoltaic modules, current transformers and inverters.

[0065] Therefore, it can be understood that the photovoltaic management system cleans the photovoltaic components, current transformers and / or inverters according to the comparison results, as shown in steps 401 to 405.

[0066] The embodiment of the present invention trains a target BIM big data photovoltaic prediction model that is jointly influenced by multiple indicator data through the indicator data under each weather type condition in the historical data and the corresponding historical photovoltaic power generation power. In the subsequent process, the indicator values ​​of the indicator data of the target photovoltaic power station under the current weather type are obtained in advance and input into the target BIM big data photovoltaic prediction model, so that the current photovoltaic power generation power of the photovoltaic power station can be obtained in advance, so that the operating equipment of the photovoltaic power station can be cleaned in advance according to the current photovoltaic power generation power, which solves the problem of management lag, avoids power generation losses and potential damage to equipment, and improves the power generation efficiency of the photovoltaic power station.

[0067] In one embodiment, the description of steps 201 to 204 is as follows:

[0068] Step 201 , constructing a data matrix by taking each weather type, the historical indicator value of the indicator data under each weather type condition, and the historical photovoltaic power generation power corresponding to the historical indicator value as a sample.

[0069] In the embodiment of the present invention, each weather type, the historical indicator value of the indicator data under each weather type condition, and the historical photovoltaic power generation corresponding to the historical indicator value are taken as a sample. At the same time, each weather type and the historical indicator value of the indicator data under each weather type condition are taken as independent variables, and the historical photovoltaic power generation corresponding to the historical indicator value is taken as the dependent variable to construct a data matrix.

[0070] In one embodiment, the weather type is represented by 0-1 coding (such as 0 for sunny, 1 for cloudy, etc.). For a sample with sunny weather type, the corresponding irradiance is I1, temperature is T1, humidity is H1, wind speed is S1, wind direction is W1, and photovoltaic power generation power is P1. Then a row of the sample in the data matrix can be represented as {0, I1, T1, H1, S1, W1, P1}.

[0071] In one embodiment, the weather type is represented by C i , where weather type C i It can be represented by integer or character code. For each weather type C i , corresponding to irradiance I i , Temperature T i 、Humidity i , Style S i 、Wind direction W i Index data, historical photovoltaic power generation power P i Therefore, the constructed data matrix X is an N*(m+1) matrix, where N is the number of samples and m is the number of independent variables. In the embodiment of the present invention, m is 6, including weather type and 5 and indicator data. Each row in the data matrix X represents a sample, and in the data matrix X, x is the number of independent variables.i1 Indicates weather type C i , with x i2 Indicates irradiance I i , with x i3 Indicates temperature T i , with x i4 Indicates humidity H i , with x i5 Indicates wind speed S i , with x i6 Indicates wind direction W i , with y i Indicates the historical photovoltaic power generation P i , therefore, the data matrix X in the embodiments of the present invention can be expressed as:

[0072]

[0073] Wherein, x ij Indicates the jth independent variable value of the ith sample (including the encoding of the weather type and the index data value), k is the total number of independent variables (the dimension of the weather type encoding + irradiance, temperature, humidity, wind speed, wind direction and other indexes), so k in the embodiments of the present invention is 6, and y i Is the historical photovoltaic power generation (dependent variable) of the ith sample.

[0074] Step 202: Take each sample in the data matrix as a cluster, and construct a cluster distance matrix based on the cluster distance between every two clusters in the data matrix.

[0075] Furthermore, each sample in the data matrix can be used as a separate cluster. At this time, the number of clusters L = N. Calculate the cluster distance between clusters through the inter-cluster distance measurement formula. The specific formula is as follows:

[0076]

[0077] Wherein, M a And M b Represent two clusters in the data matrix, and D1(M a , M b ) represents the cluster distance between cluster M a And cluster M b , p and q represent the sample vectors in cluster M a And cluster M b . In the embodiments of the present invention, the vector dimension of the data matrix is m + 1, |p - q| represents the distance measurement between vectors (such as Euclidean distance), |M a | and |M b | represent the size of the cluster (number of samples), and α represents a preset weight parameter, And Represents cluster M a and M b The main weather types in can be determined by majority voting, etc. ξ(·) represents the weather type difference function. If otherwise

[0078] Therefore, based on the inter-cluster distance measurement formula, the cluster distances between all two clusters in the data matrix are calculated to construct the cluster distance matrix H n*n .

[0079] Step 203: merge the two clusters with the smallest cluster distance in the data matrix based on the cluster distance matrix until all clusters in the data matrix are merged to obtain a target cluster tree.

[0080] Furthermore, the photovoltaic management system is based on the clustering distance matrix H n*n The two clusters with the smallest cluster distance in the data matrix are merged until all clusters in the data matrix are merged to obtain a target clustering tree, as described in steps 2031 to 2033.

[0081] Step 204 , taking the target clustering tree as the initial BIM big data photovoltaic prediction model, optimizing the initial BIM big data photovoltaic prediction model to obtain the target BIM big data photovoltaic prediction model.

[0082] Furthermore, the photovoltaic management system inputs the test data set into the initial BIM big data photovoltaic prediction model to obtain the predicted photovoltaic power generation of each test data in the test data set output by the initial BIM big data photovoltaic prediction model, and further optimizes the initial BIM big data photovoltaic prediction model according to the predicted photovoltaic power generation of each test data and the actual photovoltaic power generation of each test data to obtain the target BIM big data photovoltaic prediction model, as specifically described in steps 2041 to 2042.

[0083] The embodiment of the present invention trains a target BIM big data photovoltaic prediction model that is jointly influenced by multiple indicator data through the indicator data under each weather type condition in the historical data and the corresponding historical photovoltaic power generation power. In the subsequent process, the indicator values ​​of the indicator data of the target photovoltaic power station under the current weather type are obtained in advance and input into the target BIM big data photovoltaic prediction model, so that the current photovoltaic power generation power of the photovoltaic power station can be obtained in advance, so that the operating equipment of the photovoltaic power station can be cleaned in advance according to the current photovoltaic power generation power, which solves the problem of management lag, avoids power generation losses and potential damage to equipment, and improves the power generation efficiency of the photovoltaic power station.

[0084] In one embodiment, the description of steps 2031 to 2033 is as follows:

[0085] Step 2031, merging the two clusters with the smallest cluster distance in the data matrix based on the cluster distance matrix, until all clusters in the data matrix are merged, and an initial cluster tree is obtained;

[0086] Step 2032, determining the cluster center of the initial cluster tree, and determining the cluster allocation distance between each cluster in the initial cluster tree and the cluster center;

[0087] Step 2033: merge clusters with the same cluster allocation distance in the initial cluster tree until all clusters in the initial cluster tree are merged to obtain a target cluster tree.

[0088] Specifically, the photovoltaic management system is based on the clustering distance matrix H n*n Get the two clusters with the smallest cluster distance in the data matrix, where the two clusters with the smallest cluster distance can be expressed as M min1 and M min2 .

[0089] Furthermore, the photovoltaic management system divides the two clusters M with the smallest cluster distance in the data matrix into min1 and M min2 Merge until all clusters are merged and the initial clustering tree is obtained, which is as follows:

[0090] The photovoltaic management system divides the two clusters M with the smallest cluster distance in the data matrix into min1 and M min2 Merge and get two clusters M min1 and M min2 The new cluster M new , where the new cluster M new It can be expressed as M new =M min1 ∪M min2 , at this time the number of clusters in the data matrix L = N-1.

[0091] Furthermore, the PV management system recalculates the new cluster M new With other clusters M j For the new cluster M new With other clusters M j The cluster distance D2(M new ,M j ) is as follows:

[0092]

[0093] D max (M new ,M j )=(1-β)*γ*max[D1(Mmin1 ,M j ),D2(M min2 ,M j )];

[0094] Among them, β represents the preset weight parameter, and its value range is [0,1]; |M new | represents the new cluster M new The number of samples, therefore, |M new |=|M min1 |+|M min2 |;γ represents the preset scaling factor.

[0095] Furthermore, the photovoltaic management system uses the clustering distance D2 (M new ,M j ) Update the cluster distance matrix, and then obtain the two clusters with the smallest cluster distance in the updated cluster distance matrix for clustering, until all clusters in the data matrix are merged to obtain the initial cluster tree.

[0096] In the initial clustering tree in the embodiment of the present invention, the nodes represent clusters, and the edges between the nodes represent merging operations. As the merging proceeds, the tree gradually grows upward. Therefore, it can be understood that the new cluster is the parent node of the original two clusters, and the original two clusters are respectively the two child nodes of the new cluster.

[0097] Furthermore, the photovoltaic management system obtains the cluster center of the initial cluster tree and calculates the cluster allocation distance between each cluster in the initial cluster tree and the cluster center. The specific analysis is as follows:

[0098] The cluster center of the initial cluster tree is M center , cluster center M center It can be expressed as M center = {C, I, T, H, S, W, P}, weather type C (encoded), irradiance I, temperature T, humidity H, wind speed S, wind direction W, photovoltaic power generation power P, therefore, for the cluster center M center The components are:

[0099] For the central component Γ of weather type C C , calculate the information entropy weighted model of weather type C. In one embodiment, there are b types of weather types C, and the frequency of the i-th weather type in the cluster center M is f i , whose information entropy is U i =-f i log(f i ), that is, when f i = 0, U i = 0. Then the cluster center component Γ of weather type C C is the weather type with the minimum information entropy weighted value, that is:

[0100]

[0101] in, is the Kronecker function (when c = c i 1 if yes, 0 otherwise).

[0102] For the central component Γ of the irradiance I I , using the median based on kernel density estimation. In one embodiment, the values ​​of irradiance I are I1, I2, ..., I n , using the Gaussian kernel density estimation function Calculate the kernel density estimate for each sample as K(I j ). The samples are further sorted according to the kernel density estimate, and the irradiance value corresponding to the middle position after sorting is selected as Γ I .

[0103] For the central component Γ at temperature T T , using the harmonic mean weighted value of the temperature. In one embodiment, the temperature values ​​of temperature T are T1, T2, ..., T n , the weight of each sample is in, is the arithmetic mean of the temperature within the cluster, then the central component of temperature T

[0104] For the central component Γ of humidity H H , calculated by the fractal dimension of humidity. Consider the humidity data as a time series (assuming that the samples have a time sequence), and use the box counting method to calculate its fractal dimension D. Further select the humidity value H j , so that the vector formed by it and other humidity values ​​in the fractal dimension space best matches the overall fractal characteristics of the humidity data in the cluster, and the humidity value is taken as Γ H .

[0105] For the central component Γ of the wind speed S S , based on the centroid of the power spectral density of wind speed, calculate the power spectral density S(f) of the wind speed time series, and calculate the centroid of the power spectral density Where f is the frequency.

[0106] For the central component Γ of the wind direction W W , using the vector average method of wind direction. The wind direction is regarded as a vector, and the wind direction vector of each sample is (cos(W j ), sin(W j )), then the central component of the wind direction W

[0107] For the central component Γ of the photovoltaic power generation power P P , using rank-based weighted average. For photovoltaic power samples P1, P2, ..., P n Sort and get the rank r j , the weight is Then the central component of photovoltaic power generation power P is

[0108] In one embodiment, one of the clusters in the initial clustering tree is M1={C x ,I x ,T x ,H x ,S x ,W x ,P x}, the cluster center M of the initial clustering tree center ={Γ C ,Γ I ,Γ T ,Γ H ,Γ S ,Γ W ,Γ P}, therefore, cluster M1 and cluster center M center The clustering distribution distance between them is calculated as follows:

[0109] d(M1,M center )=μ C d C (C x ,Γ C )+μ I d I (I x ,Γ I )+μ T d T (T x ,Γ T )+

[0110] μ H d H (H x ,Γ H )+μ S d S (S x ,Γ S )+μ W d W (W x ,Γ W )+μ P d P (P x ,Γ P );

[0111] Among them, μC ,μ I ,μ T ,μ H ,μ S ,μ W ,μ P It is the weight of each component, which can be determined according to the influence of each factor on photovoltaic power generation.

[0112] For weather type distance d C (C x ,Γ C ), if C x =Γ C , then d C (C x ,Γ C )=0; otherwise, d C (C x ,Γ C )=1-sim(C x ,Γ C ), where sim(C x ,Γ C ) is the similarity between weather types, which can be determined by the semantic similarity of weather types (if relevant semantic information is available) or based on the similarity of the impact of weather types on power generation in historical data.

[0113] For the irradiance distance d I (I x ,Γ I ), the specific calculation formula is as follows:

[0114]

[0115] Among them, β is an adjustment parameter that makes the distance have different sensitivities in different irradiance ranges.

[0116] For the temperature distance d T (T x ,Γ T ), the specific calculation formula is as follows:

[0117]

[0118] in, is the average temperature, and γ is a parameter that determines the nonlinear degree of the effect of temperature on the generated power.

[0119] For humidity distance d H (H x ,Γ H ), the specific calculation formula is as follows:

[0120]

[0121] Among them, H max and H min are the maximum and minimum values ​​of humidity.

[0122] For wind speed distance d S (S x ,Γ S ), the specific calculation formula is as follows:

[0123]

[0124] Among them, S max and S min are the maximum and minimum wind speeds.

[0125] For wind direction distance d W (W x ,Γ W ), the specific calculation formula is as follows:

[0126]

[0127] For photovoltaic power generation distance d P (P x ,Γ P ), the specific calculation formula is as follows:

[0128]

[0129] Among them, P max is the maximum value of photovoltaic power generation, and α is a parameter determined by the distribution characteristics of power generation.

[0130] Furthermore, the photovoltaic management system merges clusters with the same cluster allocation distance in the initial cluster tree until all clusters in the initial cluster tree are merged to obtain a target cluster tree.

[0131] In one embodiment, the data matrix includes cluster M1, cluster M2, cluster M3, cluster M4, cluster M5, cluster M6, cluster M7, and cluster M8. The cluster distance between cluster M1 and cluster M4 is the smallest. After merging cluster M1 and cluster M4, a new cluster M9 is obtained. After clustering, the cluster distance between cluster M2 and cluster M6 is the smallest. After merging cluster M2 and cluster M6, a new cluster M7 is obtained. 10 After clustering, the cluster distance between cluster M2 and cluster M6 is the smallest. After merging cluster M2 and cluster M6, a new cluster M is obtained. 10 After clustering, the cluster distance between cluster M3 and cluster M9 is the smallest. After merging cluster M3 and cluster M9, a new cluster M is obtained. 11After clustering, the cluster distance between cluster M5 and cluster M7 is the smallest. After merging cluster M5 and cluster M7, a new cluster M is obtained. 12 After clustering, cluster M8 and cluster M 12 The cluster distance between clusters is the smallest, cluster M8 is connected to cluster M 12 After merging, we get a new cluster M 13 After clustering, cluster M 11 With clustering M 13 The cluster distance between them is the smallest, and cluster M 11 With clustering M 13 After merging, we get a new cluster M 14 After clustering, cluster M 10 With clustering M 14 Merge to get a new cluster M 15 Get the initial clustering tree, such as Figure 2 , Figure 2 It is a schematic diagram of the structure of the initial clustering tree provided by an embodiment of the present invention.

[0132] Furthermore, the cluster center of the initial cluster tree is M. Cluster M6 and cluster M 11 The cluster assignment distances of cluster M4 and cluster M5 are the same as the cluster assignment distances of cluster M, and the cluster assignment distances of cluster M3 and cluster M8 are the same as the cluster assignment distances of cluster M. Therefore, cluster M6 and cluster M 11 After merging, we get a new cluster M 17 After merging clusters M4 and M5, we get a new cluster M 18 After merging clusters M3 and M8, we get a new cluster M 19 Get the target clustering tree, such as Figure 3 , Figure 3 It is a schematic diagram of the structure of the target clustering tree provided by an embodiment of the present invention.

[0133] The embodiment of the present invention continuously optimizes the clustering tree, so that the final target BIM big data photovoltaic prediction model trained is more accurate and has stronger recognition performance, so that the current photovoltaic power generation power of the photovoltaic power station can be quickly and accurately obtained through the target BIM big data photovoltaic prediction model, so that the operating equipment of the photovoltaic power station can be cleaned in advance according to the current photovoltaic power generation power, which solves the problem of management lag, avoids power generation losses and potential damage to equipment, and improves the power generation efficiency of the photovoltaic power station.

[0134] In one embodiment, the description of step 2041 to step 2042 is as follows:

[0135] Step 2041 inputs the test data set into the initial BIM big data photovoltaic prediction model to obtain the predicted photovoltaic power generation of each test data in the test data set output by the initial BIM big data photovoltaic prediction model; Step 2042 optimizes the initial BIM big data photovoltaic prediction model based on the predicted photovoltaic power generation power and actual photovoltaic power generation power of each test data to obtain the target BIM big data photovoltaic prediction model.

[0136] Optionally, the photovoltaic management system obtains a test data set, wherein the test data set includes multiple test data, namely, the index value of the index data under each weather type condition and the actual photovoltaic power generation corresponding to the index value. Further, the photovoltaic management system inputs the index value of the index data under each weather type and each weather type condition in each test data of the test data set into the initial BIM big data photovoltaic prediction model for power prediction, and obtains the predicted photovoltaic power generation power of each test data in the test data set output by the initial BIM big data photovoltaic prediction model.

[0137] Furthermore, the photovoltaic management system optimizes the initial BIM big data photovoltaic prediction model according to the predicted photovoltaic power generation and the actual photovoltaic power generation of each test data to obtain the target BIM big data photovoltaic prediction model, specifically: based on the predicted photovoltaic power generation and the actual photovoltaic power generation of each test data, the average absolute error rate, root mean square, average variance, goodness of fit and daily accuracy of the power generation are determined; based on the average absolute error rate, root mean square, average variance, goodness of fit and daily accuracy of the power generation, the confidence of the initial BIM big data photovoltaic prediction model is determined; based on the confidence, the cluster center of the initial BIM big data photovoltaic prediction model is adjusted, and the confidence of the initial BIM big data photovoltaic prediction model after the cluster center is adjusted is cyclically obtained until the confidence is greater than or equal to the preset threshold, and the target BIM big data photovoltaic prediction model is obtained.

[0138] Specifically, the photovoltaic management system predicts the photovoltaic power generation power P according to the i-th test data. pi And the actual photovoltaic power generation power P of the i-th test data mi , calculate the absolute error value |ΔP of the i-th test data i |, therefore, the absolute error value |ΔP i | can be expressed as |ΔP i |=|P mi -P pi |. Furthermore, the PV management system calculates the absolute error value of all test data |ΔP i |Calculate the mean absolute error rate MAE of the generated power. The specific formula is as follows:

[0139]

[0140] Among them, n represents the number of test data, max(P mi ) represents the maximum value of the actual power generation, P rated Indicates the rated capacity of the target PV power station.

[0141] Furthermore, the absolute error value of the ith test data of the photovoltaic management system |ΔP i |, calculate the error square of the i-th test data (ΔP i ) 2 Therefore, the specific formula of the root mean square RMAE of the generated power is as follows:

[0142]

[0143] Furthermore, the photovoltaic management system predicts the photovoltaic power generation power P according to the i-th test data. pi And the actual photovoltaic power generation power P of the i-th test data mi , calculate the deviation e of the i-th test data i , therefore, the deviation e of the i-th test data is i Can represent e i =P mi -P pi Furthermore, the PV management system calculates the deviation e of the i-th test data. i Calculate the variance E of the i-th test data i , therefore, the variance E of the i-th test data is i It can be expressed as in, is the average deviation. Furthermore, the PV management system calculates the variance E of the i-th test data. i Calculate the average variance AV of the generated power. The specific formula is as follows:

[0144]

[0145] Among them, ρ i represents the adaptive weight, which is related to the rate of change of the local variance.

[0146] Furthermore, the photovoltaic management system calculates the actual photovoltaic power generation power P according to the i-th test data. mi And the total square sum SST is calculated by the average value of the actual photovoltaic power generation of all test data. Therefore, the total square sum SST can be expressed as in, is the average value of the actual photovoltaic power generation. Furthermore, the photovoltaic management system predicts the photovoltaic power generation P according to the i-th test data. pi And the actual photovoltaic power generation power P of the i-th test data mi, calculate the residual sum of squares SSE, so the residual sum of squares SSE can be expressed as Furthermore, the photovoltaic management system calculates the goodness of fit of the power generation according to the total sum of squares SST and the residual sum of squares SSE. The formula is as follows:

[0147]

[0148] Wherein, η represents the preset threshold, Represents the predicted photovoltaic power generation power P pi and the actual photovoltaic power generation power P pi When the difference between them is greater than the preset threshold, the penalty factor is assigned.

[0149] Furthermore, the photovoltaic management system predicts the photovoltaic power generation power P according to the i-th test data. pi , the actual photovoltaic power generation power P of the i-th test data mi As well as the rated capacity of the target photovoltaic power station, the daily accuracy DA of the power generation is calculated. The specific calculation formula for the daily accuracy DA is:

[0150]

[0151] Among them, L(t i ) represents the dynamic influencing factor that changes with the change of light duration in a day, L(·) represents the light duration ratio function, t i Indicates the time when the i-th test data is collected.

[0152] Furthermore, the PV management system determines the confidence of the initial BIM big data PV prediction model based on the average absolute error rate, root mean square, average variance, goodness of fit and daily accuracy of the generated power. The calculation formula of the confidence is as follows:

[0153]

[0154] Among them, Confidence represents confidence, θ1, θ2, θ3, θ4, θ5 represent preset allocation weights, and θ1+θ2+θ3+θ4+θ5=1.

[0155] Further, the photovoltaic management system determines whether the confidence level is greater than or equal to a preset threshold.

[0156] If not, the photovoltaic management system adjusts the cluster center of the initial BIM big data photovoltaic prediction model, and cyclically obtains the confidence of the initial BIM big data photovoltaic prediction model after the cluster center is adjusted until the confidence is greater than or equal to the preset threshold, and obtains the target BIM big data photovoltaic prediction model.

[0157] The average absolute error rate, root mean square, average variance, goodness of fit and daily accuracy of the power generation in the embodiment of the present invention continuously adjust the cluster center, so that the final target BIM big data photovoltaic prediction model trained is more accurate and has stronger recognition performance. Therefore, the current photovoltaic power generation of the photovoltaic power station can be quickly and accurately obtained through the target BIM big data photovoltaic prediction model, so that the operating equipment of the photovoltaic power station can be cleaned in advance according to the current photovoltaic power generation, which solves the problem of management lag, avoids power generation losses and potential damage to equipment, and improves the power generation efficiency of the photovoltaic power station.

[0158] In one embodiment, the description of steps 401 to 405 is as follows:

[0159] Step 401, if the power difference is within the first difference range, determine whether the operating temperature value of the photovoltaic component is within its preset operating temperature range; if not, send the first cleaning instruction to the cleaning device, so that the cleaning device cleans the photovoltaic component according to the first cleaning instruction; the cleaning device includes a vibration device, a mobile device and a water spraying device, and when cleaning, the vibration device, the mobile device and the water spraying device work together; if so, determine whether the operating temperature value of the current transformer is within its preset operating temperature range; if not, send the second cleaning instruction to the cleaning device, so that the cleaning device cleans the current transformer according to the second cleaning instruction; if so, send the third cleaning instruction to the cleaning device, so that the cleaning device cleans the inverter according to the third cleaning instruction; Step 402, if the power difference is within the second difference range, send the fourth cleaning instruction to the cleaning device, so that the cleaning device cleans the photovoltaic component, the current transformer and the inverter according to the fourth cleaning instruction.

[0160] Specifically, the photovoltaic management system compares the current photovoltaic power generation power with the preset power generation power to obtain a power comparison result, as well as the power difference between the current photovoltaic power generation power and the preset power generation power. The power comparison result is whether the current photovoltaic power generation power is greater than or less than the preset power generation power.

[0161] If the power comparison result is that the current photovoltaic power generation power is less than or equal to the preset power generation power, it means that the operating equipment of the target photovoltaic power station is not blocked at this time and does not need to be cleaned.

[0162] If the power comparison result shows that the current photovoltaic power generation is greater than the preset power generation, it means that the operating equipment of the target photovoltaic power station is blocked and needs to be cleaned. The specific analysis of cleaning the photovoltaic components, current transformers and / or inverters of the target photovoltaic power station is as follows:

[0163] In the embodiment of the present invention, the first preset difference is greater than the second preset difference. If the power difference is greater than the first preset difference, it means that the running equipment is seriously blocked. If the first difference range is less than the second difference range, the severity of the blockage in the second difference range is greater than the severity of the blockage in the first difference range. If the power difference is less than or equal to the first preset difference, it means that the running equipment is slightly blocked.

[0164] Therefore, for the case where the current photovoltaic power generation power is greater than the preset power generation power, and the power difference between the current photovoltaic power generation power and the preset power generation power is greater than the first preset difference:

[0165] Optionally, if the power difference is within the first difference range, the photovoltaic management system determines whether the operating temperature value of the photovoltaic component is within its preset operating temperature range, wherein the preset operating temperature range of the photovoltaic component is adjusted according to the season and working hours.

[0166] Furthermore, if it is determined that the operating temperature value of the photovoltaic module is outside its preset operating temperature range, it means that the photovoltaic module of the target photovoltaic power station is blocked, so the photovoltaic management system sends a first cleaning instruction to the cleaning device. After receiving the first cleaning instruction, the cleaning device cleans the photovoltaic module according to the first cleaning instruction, wherein the cleaning device includes a vibration device, a mobile device and a water spray device, and when cleaning, the vibration device, the mobile device and the water spray device work together.

[0167] Furthermore, if it is determined that the operating temperature value of the photovoltaic component is within its preset operating temperature range, it means that there is no blockage in the photovoltaic components of the target photovoltaic power station. Therefore, the photovoltaic management system determines whether the operating temperature value of the current transformer is within its preset operating temperature range, or whether the operating temperature value of the inverter is within its preset operating temperature range.

[0168] Furthermore, if it is determined that the operating temperature value of the current transformer is outside its preset operating temperature range, it indicates that the current transformer of the target photovoltaic power station is blocked, so the photovoltaic management system sends a second cleaning instruction to the cleaning device. After receiving the second cleaning instruction, the cleaning device cleans the current transformer according to the second cleaning instruction. Similarly, the vibration device, the mobile device and the water spray device perform cleaning together.

[0169] Furthermore, if it is determined that the operating temperature value of the current transformer is within its preset operating temperature range, it means that the current transformer of the target photovoltaic power station is not blocked and the inverter of the target photovoltaic power station is blocked. Therefore, the photovoltaic management system sends the third cleaning instruction to the cleaning device. After receiving the third cleaning instruction, the cleaning device cleans the inverter according to the third cleaning instruction. Similarly, the vibration device, the mobile device and the water spraying device perform cleaning together.

[0170] Optionally, if the power difference is within the second difference range, the photovoltaic management system sends a fourth cleaning instruction to the cleaning device. After receiving the fourth cleaning instruction, the cleaning device cleans the photovoltaic components, current transformers and inverters at the same time according to the fourth cleaning instruction. Similarly, the vibration device, mobile device and water spray device are cleaned together.

[0171] For the case where the current photovoltaic power generation power is greater than the preset power generation power, and the power difference between the current photovoltaic power generation power and the preset power generation power is greater than or equal to the first preset difference:

[0172] In step 403, a start instruction is sent to the camera device corresponding to the target photovoltaic power station to obtain a first top image of the photovoltaic component, a second top image of the current transformer and a third top image of the inverter, which are collected by the camera device based on the start instruction; in step 404, image detection is performed on the first top image, the second top image and the third top image to obtain a target top image with obstructions; in step 405, a vibration instruction is sent to the cleaning device based on the target top image, so that the cleaning device removes obstructions on the top of the photovoltaic component, the current transformer or the inverter based on the vibration instruction.

[0173] Optionally, the photovoltaic management system sends a start instruction to the camera device corresponding to the target photovoltaic power station. After receiving the start instruction, the camera device turns on the camera according to the start instruction to collect the first top image of the photovoltaic component, the second top image of the current transformer, and the third top image of the inverter, and transmits the collected first top image of the photovoltaic component, the second top image of the current transformer, and the third top image of the inverter back to the photovoltaic management system.

[0174] Further, image detection is performed on the first top image, the second top image, and the third top image to obtain a target top image with an occluder, so as to perform image detection on the first top image. Specifically, as follows:

[0175] Optionally, the photovoltaic management system divides the first top image into M sub-areas, for example, M can be 16, 25, 36, etc., and the image area of ​​the i-th sub-area is s i .

[0176] Furthermore, the photovoltaic management system obtains the total number of pixels n in the i-th sub-area i , according to the total number of pixels n in the i-th sub-region i and the image area is s i Calculate the pixel density d of the i-th sub-area i , therefore, the pixel density d i Can represent d i =n i / si .

[0177] Furthermore, the photovoltaic management system obtains the grayscale value of each pixel in the i-th sub-region and converts the grayscale value g of each pixel in the i-th sub-region into j Sum and divide by the total number of pixels to get the average grayscale value G of the i-th sub-region i , therefore, the average gray value G i It can be expressed as

[0178] Furthermore, the photovoltaic management system calculates the grayscale value of each pixel in the ith sub-region and the average grayscale value G of the ith sub-region. i The square of the difference is added to get the calculation result z of the i-th sub-region i , so the calculated result z i It can be expressed as

[0179] Furthermore, the photovoltaic management system calculates the value of z in the ith sub-area. i And the total number of pixels n i Get the grayscale standard deviation σ of the i-th sub-region i , therefore, the grayscale standard deviation σ i Can be expressed

[0180] Furthermore, the photovoltaic management system calculates the characteristic index R1 related to the sub-area, and the specific formula is as follows:

[0181]

[0182] Furthermore, the photovoltaic management system calculates the globally relevant characteristic index R2, and the specific formula is as follows:

[0183]

[0184] Among them, d all represents the average pixel density of the first top image, n all Indicates the total number of pixels in the first top image, s all represents the area of ​​the first top image, σ all represents the average grayscale standard deviation of the first top image, G all represents the average gray value of the first top image, and v represents the number of channels of the first top image.

[0185] Furthermore, the photovoltaic management system calculates the comprehensive index of the first top image according to the sub-region-related characteristic index R1 and the global-related characteristic index R2. The specific formula is:

[0186] R = τ*R1+(1-τ)*R2;

[0187] Wherein, τ represents the preset weight coefficient.

[0188] Further, if it is determined that the comprehensive index of the first top image is greater than a preset index threshold, the photovoltaic management system determines that there is an obstruction in the first top image.

[0189] Furthermore, the photovoltaic management system sends a vibration command to the cleaning device according to the target top image. After receiving the vibration command, the cleaning device triggers the vibration module to work according to the vibration command to remove the obstruction on the top of the photovoltaic module, current transformer or inverter.

[0190] Furthermore, after the photovoltaic management system sends a vibration instruction to the cleaning device to clean the obstructions on the top of the photovoltaic module, current transformer or inverter, a re-collection instruction is sent to the camera device, and the camera device re-collects the first top image of the photovoltaic module, the second top image of the current transformer and the third top image of the inverter according to the re-collection instruction. The photovoltaic management system detects the image to determine whether the obstruction has been removed. If not, a manual cleaning instruction is sent to the external terminal.

[0191] The embodiment of the present invention cleans the operating equipment of the target photovoltaic power station in different ways according to the current photovoltaic power generation power, thereby improving the management flexibility of the operating equipment. At the same time, different operating equipment is cleaned according to the power difference range, thereby improving the management accuracy of the operating equipment.

[0192] Furthermore, the intelligent photovoltaic management system based on the BIM big data model provided by the present invention is described below. The intelligent photovoltaic management system based on the BIM big data model described below and the intelligent photovoltaic management method based on the BIM big data model described above can be referenced to each other.

[0193] Optional, see Figure 4 , Figure 4 It is a structural schematic diagram of the intelligent photovoltaic management system based on the BIM big data model provided by the present invention. The intelligent photovoltaic management system based on the BIM big data model includes a data analysis module 410, a model training module 420, a power prediction module 430 and an equipment management module 440.

[0194] The data analysis module 410 is used to analyze the historical data of the target photovoltaic power station to obtain the historical index values ​​of the index data of the target photovoltaic power station under each weather type condition and the historical photovoltaic power generation power corresponding to the historical index values; the index data includes at least one of irradiance, temperature, humidity, wind speed and wind direction;

[0195] A model training module 420 is used to perform model training based on the historical indicator values ​​of the indicator data under each weather type condition and the historical photovoltaic power generation power corresponding to the historical indicator values ​​to obtain a target BIM big data photovoltaic prediction model;

[0196] The power prediction module 430 is used to input the target index value of the index data of the target photovoltaic power station under the current weather type into the target BIM big data photovoltaic prediction model to obtain the current photovoltaic power generation power output by the target BIM big data photovoltaic prediction model;

[0197] The equipment management module 440 is used to clean the operating equipment of the target photovoltaic power station based on the current photovoltaic power generation power; the operating equipment includes at least one of a photovoltaic module, a current transformer and an inverter.

[0198] The embodiment of the present invention trains a target BIM big data photovoltaic prediction model that is jointly influenced by multiple indicator data through the indicator data under each weather type condition in the historical data and the corresponding historical photovoltaic power generation power. In the subsequent process, the indicator values ​​of the indicator data of the target photovoltaic power station under the current weather type are obtained in advance and input into the target BIM big data photovoltaic prediction model, so that the current photovoltaic power generation power of the photovoltaic power station can be obtained in advance, so that the operating equipment of the photovoltaic power station can be cleaned in advance according to the current photovoltaic power generation power, which solves the problem of management lag, avoids power generation losses and potential damage to equipment, and improves the power generation efficiency of the photovoltaic power station.

[0199] See also Figure 5 , Figure 5 FIG. 1 is an embodiment diagram of an electronic device provided by an embodiment of the present invention. Figure 5 As shown, an embodiment of the present invention provides an electronic device 500, including a memory 510, a processor 520, and a computer program 511 stored in the memory 510 and executable on the processor 520. When the processor 520 executes the computer program 511, the following steps are implemented:

[0200] Perform data analysis on the historical data of the target photovoltaic power station to obtain the historical index values ​​of the index data of the target photovoltaic power station under each weather type condition and the historical photovoltaic power generation power corresponding to the historical index values; the index data includes at least one of irradiance, temperature, humidity, wind speed and wind direction;

[0201] Model training is performed based on the historical indicator values ​​of the indicator data under each weather type condition and the historical photovoltaic power generation power corresponding to the historical indicator values, and the target BIM big data photovoltaic prediction model is obtained;

[0202] Input the target index value of the index data of the target photovoltaic power station under the current weather type into the target BIM big data photovoltaic prediction model to obtain the current photovoltaic power generation power output by the target BIM big data photovoltaic prediction model;

[0203] The operating equipment of the target photovoltaic power station is cleaned based on the current photovoltaic power generation power; the operating equipment includes at least one of a photovoltaic component, a current transformer and an inverter.

[0204] See also Figure 6 , Figure 6 Detailed description of an embodiment of a computer-readable storage medium provided by an embodiment of the present invention. Figure 6 As shown, this embodiment provides a computer-readable storage medium 600, on which a computer program 511 is stored. When the computer program 511 is executed by a processor, the following steps are implemented:

[0205] Perform data analysis on the historical data of the target photovoltaic power station to obtain the historical index values ​​of the index data of the target photovoltaic power station under each weather type condition and the historical photovoltaic power generation power corresponding to the historical index values; the index data includes at least one of irradiance, temperature, humidity, wind speed and wind direction;

[0206] Model training is performed based on the historical indicator values ​​of the indicator data under each weather type condition and the historical photovoltaic power generation power corresponding to the historical indicator values, and the target BIM big data photovoltaic prediction model is obtained;

[0207] Input the target index value of the index data of the target photovoltaic power station under the current weather type into the target BIM big data photovoltaic prediction model to obtain the current photovoltaic power generation power output by the target BIM big data photovoltaic prediction model;

[0208] The operating equipment of the target photovoltaic power station is cleaned based on the current photovoltaic power generation power; the operating equipment includes at least one of a photovoltaic component, a current transformer and an inverter.

[0209] On the other hand, the present invention further provides a computer program product, which includes a computer program. The computer program can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the intelligent photovoltaic management method based on the BIM big data model provided by the above methods, and the method includes:

[0210] Perform data analysis on the historical data of the target photovoltaic power station to obtain the historical index values ​​of the index data of the target photovoltaic power station under each weather type condition and the historical photovoltaic power generation power corresponding to the historical index values; the index data includes at least one of irradiance, temperature, humidity, wind speed and wind direction;

[0211] Model training is performed based on the historical indicator values ​​of the indicator data under each weather type condition and the historical photovoltaic power generation power corresponding to the historical indicator values, and the target BIM big data photovoltaic prediction model is obtained;

[0212] Input the target index value of the index data of the target photovoltaic power station under the current weather type into the target BIM big data photovoltaic prediction model to obtain the current photovoltaic power generation power output by the target BIM big data photovoltaic prediction model;

[0213] The operating equipment of the target photovoltaic power station is cleaned based on the current photovoltaic power generation power; the operating equipment includes at least one of a photovoltaic component, a current transformer and an inverter.

[0214] The system embodiment described above is merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art may understand and implement it without creative work.

[0215] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0216] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. An intelligent photovoltaic management method based on BIM big data model, characterized in that: include: Performing data analysis on the historical data of the target photovoltaic power station to obtain the historical index values ​​of the index data of the target photovoltaic power station under each weather type condition and the historical photovoltaic power generation power corresponding to the historical index values; the index data includes at least one of irradiance, temperature, humidity, wind speed and wind direction; Model training is performed based on the historical indicator values ​​of the indicator data under each weather type condition and the historical photovoltaic power generation power corresponding to the historical indicator values, and the target BIM big data photovoltaic prediction model is obtained; Inputting the target index value of the index data of the target photovoltaic power station under the current weather type into the target BIM big data photovoltaic prediction model to obtain the current photovoltaic power generation power output by the target BIM big data photovoltaic prediction model; Based on the current photovoltaic power generation power, equipment cleaning is performed on operating equipment of the target photovoltaic power station; the operating equipment includes at least one of a photovoltaic component, a current transformer and an inverter.

2. The intelligent photovoltaic management method based on the BIM big data model according to claim 1 is characterized in that: The model training is performed based on the historical indicator values ​​of the indicator data under each weather type condition and the historical photovoltaic power generation power corresponding to the historical indicator values ​​to obtain the target BIM big data photovoltaic prediction model, including: Taking each weather type, the historical indicator value of the indicator data under each weather type condition, and the historical photovoltaic power generation power corresponding to the historical indicator value as a sample, a data matrix is ​​constructed; wherein, in the data matrix, each weather type and the historical indicator value of the indicator data under each weather type condition are independent variables, and the historical photovoltaic power generation power corresponding to the historical indicator value is the dependent variable; Taking each sample in the data matrix as a cluster, and constructing a cluster distance matrix based on the cluster distance between every two clusters in the data matrix; Based on the cluster distance matrix, two clusters with the smallest cluster distance in the data matrix are merged until all clusters in the data matrix are merged to obtain a target cluster tree; in the target cluster tree, nodes represent clusters, edges between nodes represent cluster merging operations, the new cluster is the parent node of the two original clusters, and the two original clusters are respectively the two child nodes of the new cluster; The target clustering tree is used as an initial BIM big data photovoltaic prediction model, and the initial BIM big data photovoltaic prediction model is optimized to obtain the target BIM big data photovoltaic prediction model.

3. The intelligent photovoltaic management method based on the BIM big data model according to claim 2 is characterized in that: The step of merging two clusters with the smallest cluster distance in the data matrix based on the cluster distance matrix until all clusters in the data matrix are merged to obtain a target cluster tree includes: Based on the cluster distance matrix, two clusters with the smallest cluster distance in the data matrix are merged until all clusters in the data matrix are merged to obtain an initial cluster tree; Determining a cluster center of the initial cluster tree, and determining a cluster allocation distance between each cluster in the initial cluster tree and the cluster center; The clusters with the same cluster allocation distance in the initial clustering tree are merged until all clusters in the initial clustering tree are merged to obtain the target clustering tree.

4. The intelligent photovoltaic management method based on the BIM big data model according to claim 2 is characterized in that: The optimizing the initial BIM big data photovoltaic prediction model to obtain the target BIM big data photovoltaic prediction model includes: Input the test data set into the initial BIM big data photovoltaic prediction model to obtain the predicted photovoltaic power generation of each test data in the test data set output by the initial BIM big data photovoltaic prediction model; Based on the predicted photovoltaic power generation and the actual photovoltaic power generation of each test data, the initial BIM big data photovoltaic prediction model is optimized to obtain the target BIM big data photovoltaic prediction model.

5. The intelligent photovoltaic management method based on the BIM big data model according to claim 4 is characterized in that: The predicted photovoltaic power generation and the actual photovoltaic power generation of each test data are used to optimize the initial BIM big data photovoltaic prediction model to obtain the target BIM big data photovoltaic prediction model, including: Based on the predicted PV power and the actual PV power for each test data, the mean absolute error rate, root mean square, average variance, goodness of fit and daily accuracy of the power are determined; Determine the confidence of the initial BIM big data photovoltaic prediction model based on the mean absolute error rate, root mean square, average variance, goodness of fit and daily accuracy of the generated power; The cluster center of the initial BIM big data photovoltaic prediction model is adjusted based on the confidence level, and the confidence level of the initial BIM big data photovoltaic prediction model after the cluster center adjustment is cyclically obtained until the confidence level is greater than or equal to a preset threshold, so as to obtain the target BIM big data photovoltaic prediction model.

6. The intelligent photovoltaic management method based on the BIM big data model according to any one of claims 1 to 5, characterized in that: The cleaning of the operating equipment of the target photovoltaic power station based on the current photovoltaic power generation power includes: For the case where the current photovoltaic power generation power is greater than the preset power generation power, and the power difference between the current photovoltaic power generation power and the preset power generation power is greater than the first preset difference: If the power difference is within the first difference range, it is determined whether the operating temperature value of the photovoltaic component is within its preset operating temperature range; if not, a first cleaning instruction is sent to the cleaning device, so that the cleaning device cleans the photovoltaic component according to the first cleaning instruction; the cleaning device includes a vibration device, a mobile device and a water spraying device, and when cleaning, the vibration device, the mobile device and the water spraying device work together; if yes, it is determined whether the operating temperature value of the current transformer is within its preset operating temperature range; if not, a second cleaning instruction is sent to the cleaning device, so that the cleaning device cleans the current transformer according to the second cleaning instruction; if yes, a third cleaning instruction is sent to the cleaning device, so that the cleaning device cleans the inverter according to the third cleaning instruction; If the power difference is within the second difference range, a fourth cleaning instruction is sent to the cleaning device, so that the cleaning device cleans the photovoltaic component, the current transformer and the inverter according to the fourth cleaning instruction.

7. The intelligent photovoltaic management method based on the BIM big data model according to claim 6 is characterized in that: The method further comprises: For the case where the current photovoltaic power generation power is greater than the preset power generation power, and the power difference between the current photovoltaic power generation power and the preset power generation power is greater than or equal to the first preset difference: Sending a start instruction to a camera device corresponding to the target photovoltaic power station to obtain a first top image of the photovoltaic assembly, a second top image of the current transformer, and a third top image of the inverter acquired by the camera device based on the start instruction; Performing image detection on the first top image, the second top image, and the third top image to obtain a target top image with an occluder; A vibration instruction is sent to the cleaning device based on the target top image, so that the cleaning device removes obstructions on the top of the photovoltaic component, the current transformer or the inverter based on the vibration instruction.

8. An intelligent photovoltaic management system based on BIM big data model, characterized in that: Applicable to the intelligent photovoltaic management method based on the BIM big data model as described in any one of claims 1 to 7; the intelligent photovoltaic management system based on the BIM big data model includes a data analysis module, a model training module, a power prediction module and an equipment management module; A data analysis module is used to analyze the historical data of the target photovoltaic power station to obtain the historical index values ​​of the index data of the target photovoltaic power station under each weather type condition and the historical photovoltaic power generation power corresponding to the historical index values; the index data includes at least one of irradiance, temperature, humidity, wind speed and wind direction; A model training module is used to perform model training based on the historical indicator values ​​of the indicator data under each weather type condition and the historical photovoltaic power generation power corresponding to the historical indicator values, so as to obtain the target BIM big data photovoltaic prediction model; A power prediction module, used to input the target index value of the index data of the target photovoltaic power station under the current weather type into the target BIM big data photovoltaic prediction model, and obtain the current photovoltaic power generation power output by the target BIM big data photovoltaic prediction model; The equipment management module is used to clean the operating equipment of the target photovoltaic power station based on the current photovoltaic power generation power; the operating equipment includes at least one of a photovoltaic module, a current transformer and an inverter.

9. An electronic device, comprising: Memory for storing computer software programs; A processor, used to read and execute the computer software program, characterized in that when the processor executes the computer software program, it implements the intelligent photovoltaic management method based on the BIM big data model as described in any one of claims 1 to 7.

10. A non-transitory computer-readable storage medium having a computer software program stored therein, characterized in that: When the computer software program is executed by a processor, the intelligent photovoltaic management method based on the BIM big data model as described in any one of claims 1 to 7 is implemented.

Images