CN115796338A - Photovoltaic power generation power prediction model construction and photovoltaic power generation power prediction method - Google Patents
Photovoltaic power generation power prediction model construction and photovoltaic power generation power prediction method Download PDFInfo
- Publication number
- CN115796338A CN115796338A CN202211407351.5A CN202211407351A CN115796338A CN 115796338 A CN115796338 A CN 115796338A CN 202211407351 A CN202211407351 A CN 202211407351A CN 115796338 A CN115796338 A CN 115796338A
- Authority
- CN
- China
- Prior art keywords
- monitoring
- power generation
- data
- photovoltaic power
- vector
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
- Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
- Y04S10/00—Systems supporting electrical power generation, transmission or distribution
- Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
Landscapes
- Photovoltaic Devices (AREA)
Abstract
The invention relates to the technical field of photovoltaic power generation, in particular to a photovoltaic power generation power prediction model construction and a photovoltaic power generation power prediction method, wherein the method comprises the steps of firstly obtaining a plurality of sample vectors; then, removing data of corresponding irrelevant monitoring classes in the plurality of sample vectors according to the relevance of the generated power and the plurality of monitoring classes; and finally, adjusting a plurality of parameters of the photovoltaic power generation initial model according to the plurality of sample vectors to obtain a photovoltaic power generation power prediction model. According to the embodiment of the invention, when the generated power is predicted, some data irrelevant to the generated power are removed according to the generated power, so that the complexity of the model is reduced, and the calculation amount of model construction is reduced. The embodiment of the invention also sets the input layer of the LSTM node according to the relevance of the generated power and the long-term data, and the influence of the long-term data is considered, so that the model is closer to the physical reality and is more accurate to construct, thereby ensuring the prediction precision of the model.
Description
Technical Field
The invention relates to the technical field of photovoltaic power generation, in particular to a photovoltaic power generation power prediction model construction method and a photovoltaic power generation power prediction method.
Background
With the large-scale grid connection of low-voltage distributed photovoltaic, a power distribution main body is more complex and diversified.
Meanwhile, the double fluctuation of the supply side and the demand side can cause the phenomenon of time-space mismatching of the distributed photovoltaic output and the electric load, the local consumption of the distributed photovoltaic output is difficult, and the operation and management problems generated by the local consumption are increasingly obvious.
In order to solve the problems of the low-voltage distribution network in the aspects of balance of supply and demand, optimal operation, power grid safety and the like, the prior art researches photovoltaic power generation modes from multiple aspects, wherein the prediction of photovoltaic power generation has important significance for the operation control of photovoltaic power generation.
As we know, photovoltaic power generation is affected by multiple factors, some of which do not have conditions for direct access. For example, weather is indispensable to influence of power generation power, and considering that the output of distributed photovoltaic has strong fluctuation characteristics under the influence of weather factors, the distributed photovoltaic is not provided with a special weather monitoring station and lacks of weather actual measurement data, and it is difficult to obtain an ideal prediction effect by directly establishing a power generation power prediction model according to historical power.
Therefore, it is necessary to specify factors affecting the photovoltaic power generation power and estimate some of the factors that are not easily obtained, thereby reducing the effect of affecting the power generation power prediction due to the small number of factor data.
Based on the above, a photovoltaic power generation power prediction model construction method needs to be developed and designed.
Disclosure of Invention
The embodiment of the invention provides a photovoltaic power generation power prediction model construction method and a photovoltaic power generation power prediction method, which are used for solving the problem of inaccurate photovoltaic power generation power prediction in the prior art.
In a first aspect, an embodiment of the present invention provides a photovoltaic power generation power prediction model construction method, including:
acquiring a plurality of sample vectors, wherein each sample vector comprises generated power data of a target photovoltaic power generation device and a plurality of monitoring data acquired based on a plurality of monitoring classes, and the plurality of monitoring data and the generated power data are acquired based on the same time node;
removing data corresponding to irrelevant monitoring classes in the plurality of sample vectors according to the relevance between the generated power and the plurality of monitoring classes, wherein the relevance between the irrelevant monitoring classes and the generated power is lower than a threshold value;
and adjusting a plurality of parameters of the photovoltaic power generation initial model according to the plurality of sample vectors to obtain a photovoltaic power generation power prediction model, wherein the photovoltaic power generation initial model is constructed based on an artificial neural network.
In a possible implementation manner, the removing, according to the correlation between the generated power and the multiple monitoring classes, data of corresponding unrelated monitoring classes in the multiple sample vectors includes:
sorting the plurality of sample vectors in a predetermined order;
according to the plurality of sample vectors, respectively obtaining a power generation power vector and a plurality of monitoring class vectors, wherein a plurality of elements of the power generation power vector are obtained based on power generation data of the plurality of sample vectors, and a plurality of elements of the monitoring class vectors are obtained based on monitoring data of the same monitoring class in the plurality of sample vectors;
obtaining a plurality of first correlation coefficients according to the generated power vector and the plurality of monitoring class vectors, wherein the plurality of first correlation coefficients correspond to the plurality of monitoring classes, and the first correlation coefficients represent the correlation between the monitoring classes and the generated power;
taking the monitoring class with the absolute value of the first correlation coefficient lower than the threshold value as the irrelevant monitoring class;
and removing the monitoring data of the corresponding irrelevant monitoring class in the plurality of sample vectors.
In one possible implementation manner, the obtaining a plurality of first correlation coefficients according to the generated power vector and the plurality of monitoring class vectors includes:
obtaining the plurality of first correlation coefficients according to a first formula, the generated power vector and the plurality of monitoring class vectors, wherein the first formula is as follows:
in the formula, P i A first correlation coefficient between the ith monitoring class and the generated power, N is the total number of generated power vector elements, a n Is the n-th element of the generated power vector, c in Is the nth element of the ith difference-like vector, b in Is the nth element of the ith monitoring class.
In one possible implementation, the initial model of photovoltaic power generation includes: the monitoring system comprises an input layer, a hidden layer and an output node, wherein the number of nodes of the input layer is the same as that of a plurality of monitoring data in a sample vector, the hidden layer comprises a plurality of fully-connected intermediate nodes, the input of the hidden layer is connected with the input layer, and the output node is connected with the output of the hidden layer;
the nodes corresponding to the memory classes in the input layer are constructed based on an LSTM model, wherein the memory classes are monitoring classes of long-term data influencing the generated power;
adjusting a plurality of parameters of the photovoltaic power generation initial model according to the plurality of sample vectors to obtain a photovoltaic power generation power prediction model, including:
arranging the plurality of sample vectors according to the time sequence;
a sample input step: inputting a plurality of monitoring data of each sample vector in the plurality of sample vectors into the photovoltaic power generation initial model to obtain a plurality of model outputs;
obtaining residual errors of the model outputs according to the model outputs and the generated power data of the sample vectors;
if the residual error is larger than the threshold value, adjusting a plurality of parameters of the initial photovoltaic power generation model according to the residual error and a gradient descent algorithm, and skipping to the sample input step;
and if not, taking the photovoltaic power generation initial model as the photovoltaic power generation power prediction model.
In one possible implementation, the memory class is determined according to the following steps:
arranging the plurality of sample vectors according to the time sequence;
respectively obtaining a power generation power vector and a plurality of monitoring class vectors according to the plurality of sample vectors, wherein a plurality of elements of the power generation power vector are obtained based on power generation data of the plurality of sample vectors, and a plurality of elements of the monitoring class vectors are obtained based on monitoring data of the same monitoring class in the plurality of sample vectors;
obtaining a plurality of accumulated vectors according to the plurality of monitoring class vectors and a second formula, wherein the second formula is as follows:
in the formula, ar i (n + j) is the ith monitorN + j element of class, b in Is the nth element of the ith monitoring class, and j is the accumulation order;
obtaining a plurality of second correlation coefficients according to the generated power vector and the accumulated vectors, wherein the second correlation coefficients correspond to the monitoring classes, and the second correlation coefficients represent the correlation between the long-term data of the monitoring classes and the generated power;
and taking the monitoring class with the absolute value of the second correlation coefficient higher than the threshold value as the memory class.
In a second aspect, an embodiment of the present invention provides a photovoltaic power generation power prediction method, including:
obtaining a photovoltaic power generation power prediction model constructed by the photovoltaic power generation power prediction model construction method according to the first aspect or any possible implementation manner of the first aspect;
acquiring a plurality of current data and a plurality of historical data sets according to the plurality of monitoring classes, wherein the historical data sets comprise a plurality of historical data, and the historical data is the historical monitoring data corresponding to the monitoring class lacking the current monitoring data;
generating a plurality of gray data corresponding to the current moment according to the plurality of historical data sets;
and inputting the current data and the grey data into the photovoltaic power generation power prediction model to obtain the predicted power generation power.
In one possible implementation, the generating a plurality of gray data corresponding to a current time from the plurality of historical data sets includes:
for each historical data set, the following steps are performed:
obtaining a sum vector according to a plurality of historical data of a historical data set and a third formula, wherein the third formula is as follows:
in the formula, h 1 (k) Is the kth element of the sum vector, K being in the historical datasetTotal number of elements, h 0 (k) Is the mth element of the historical data set;
obtaining a mean vector according to the sum vector and a fourth formula, wherein the fourth formula is as follows:
in the formula, av 1 (k) The kth element of the mean vector;
determining a first weight and a second weight according to the sum vector, the mean vector and a fifth formula, wherein the fifth formula is as follows:
in the formula, A 1 First weight, A 2 Is a second weight, h 0 (K) The Kth element of the historical data set;
determining gray data according to the first weight, the second weight, the historical data set, and a sixth formula, wherein the sixth formula is:
wherein e is a natural constant, h 0 (K + 1) is gray data.
In a third aspect, an embodiment of the present invention provides a photovoltaic power generation prediction model building apparatus, configured to implement a photovoltaic power generation prediction model building method according to the first aspect or any possible implementation manner of the first aspect, where the photovoltaic power generation prediction model building apparatus includes:
the system comprises a sample vector acquisition module, a data processing module and a data processing module, wherein the sample vector acquisition module is used for acquiring a plurality of sample vectors, the sample vectors comprise generated power data of target photovoltaic power generation equipment and a plurality of monitoring data acquired based on a plurality of monitoring classes, and the plurality of monitoring data and the generated power data are acquired based on the same time node;
the data removing module is used for removing data corresponding to irrelevant monitoring classes in the sample vectors according to the relevance between the generated power and the monitoring classes, wherein the relevance between the irrelevant monitoring classes and the generated power is lower than a threshold value;
and the number of the first and second groups,
and the model building module is used for adjusting a plurality of parameters of the photovoltaic power generation initial model according to the plurality of sample vectors to obtain a photovoltaic power generation power prediction model, wherein the photovoltaic power generation initial model is built on the basis of an artificial neural network.
In a fourth aspect, an embodiment of the present invention provides a terminal, including a memory and a processor, where the memory stores a computer program operable on the processor, and the processor executes the computer program to implement the steps of the method according to the first aspect, any possible implementation manner of the first aspect, the second aspect, or any possible implementation manner of the second aspect.
In a fifth aspect, the present invention provides a computer-readable storage medium, which stores a computer program that, when executed by a processor, implements the steps of the method as described above in the first aspect, any of the possible implementations of the first aspect, the second aspect, or any of the possible implementations of the second aspect.
Compared with the prior art, the implementation mode of the invention has the following beneficial effects:
the embodiment of the invention discloses a photovoltaic power generation power prediction model construction method, which comprises the steps of firstly obtaining a plurality of sample vectors, wherein the sample vectors comprise power generation power data of target photovoltaic power generation equipment and a plurality of monitoring data obtained based on a plurality of monitoring classes, and the plurality of monitoring data and the power generation power data are obtained based on the same time node; then, removing data corresponding to irrelevant monitoring classes in the plurality of sample vectors according to the relevance between the generated power and the plurality of monitoring classes, wherein the relevance between the irrelevant monitoring classes and the generated power is lower than a threshold value; and finally, adjusting a plurality of parameters of the photovoltaic power generation initial model according to the plurality of sample vectors to obtain a photovoltaic power generation power prediction model, wherein the photovoltaic power generation initial model is constructed based on an artificial neural network. According to the embodiment of the invention, when the generated power is predicted, some data irrelevant to the generated power are removed according to the generated power, so that the complexity of the model is reduced, and the calculation amount of model construction is reduced. The embodiment of the invention also sets the input layer of the LSTM node according to the relevance of the generated power and the long-term data, and the influence of the long-term data is considered, so that the model is closer to the physical reality and is more accurate to construct, thereby ensuring the prediction precision of the model.
In the aspect of photovoltaic power generation power prediction, firstly, a plurality of current data and a plurality of historical data sets are obtained according to a plurality of monitoring classes, wherein the historical data sets comprise a plurality of historical data, and the historical data is the historical monitoring data corresponding to the monitoring class lacking the current monitoring data; then, generating a plurality of gray data corresponding to the current moment according to the plurality of historical data sets; and finally, inputting the current data and the grey data into the photovoltaic power generation power prediction model to obtain the predicted power generation power. When the prediction model is used, some data which cannot be directly obtained are complemented, so that the number of the construction of monitoring points can be reduced, the consumption of resources is reduced, and the accuracy of the predicted data is ensured.
Drawings
In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings used in the embodiments or the description of the prior art will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art that other drawings can be obtained according to the drawings without inventive labor.
FIG. 1 is a flow chart of a photovoltaic power generation power prediction model construction method provided by an embodiment of the invention;
FIG. 2 is a functional block diagram of an initial model of photovoltaic power generation provided by an embodiment of the invention;
FIG. 3 is a functional block diagram of a photovoltaic power generation power prediction model construction device provided by the embodiment of the invention;
fig. 4 is a functional block diagram of a terminal according to an embodiment of the present invention.
Detailed Description
In the following description, for purposes of explanation and not limitation, specific details are set forth, such as particular system structures, techniques, etc. in order to provide a thorough understanding of the embodiments of the invention. It will be apparent, however, to one skilled in the art that the present invention may be practiced in other embodiments that depart from these specific details. In other instances, detailed descriptions of well-known systems, devices, and methods are omitted so as not to obscure the description of the present invention with unnecessary detail.
In order to make the objects, technical solutions and advantages of the present invention more apparent, the following description is made with reference to the accompanying drawings.
The following is a detailed description of the embodiments of the present invention, which is implemented on the premise of the technical solution of the present invention, and a detailed implementation manner and a specific operation process are given, but the scope of the present invention is not limited to the following embodiments.
Fig. 1 is a flowchart of a photovoltaic power generation power prediction model construction method according to an embodiment of the present invention.
In a first aspect, as shown in fig. 1, an implementation flowchart of a photovoltaic power generation power prediction model construction method provided by an embodiment of the present invention is shown, and details are as follows:
in step 101, a plurality of sample vectors are obtained, wherein a sample vector includes generated power data of a target photovoltaic power generation device and a plurality of monitoring data obtained based on a plurality of monitoring classes, and the plurality of monitoring data and the generated power data are obtained based on the same time node.
In step 102, removing data corresponding to an irrelevant monitoring class from the plurality of sample vectors according to the relevance between the generated power and the plurality of monitoring classes, wherein the relevance between the irrelevant monitoring class and the generated power is lower than a threshold value.
In some embodiments, the step 102 comprises:
sorting the plurality of sample vectors in a predetermined order;
according to the plurality of sample vectors, respectively obtaining a power generation power vector and a plurality of monitoring class vectors, wherein a plurality of elements of the power generation power vector are obtained based on power generation data of the plurality of sample vectors, and a plurality of elements of the monitoring class vectors are obtained based on monitoring data of the same monitoring class in the plurality of sample vectors;
obtaining a plurality of first correlation coefficients according to the generated power vector and the plurality of monitoring class vectors, wherein the plurality of first correlation coefficients correspond to the plurality of monitoring classes, and the first correlation coefficients represent the correlation between the monitoring classes and the generated power;
taking the monitoring class with the absolute value of the first correlation coefficient lower than the threshold value as the irrelevant monitoring class;
and removing the monitoring data of the corresponding irrelevant monitoring class in the plurality of sample vectors.
In some embodiments, said obtaining a plurality of first correlation coefficients from said generated power vector and said plurality of monitoring class vectors comprises:
obtaining the plurality of first correlation coefficients according to a first formula, the generated power vector and the plurality of monitoring class vectors, wherein the first formula is as follows:
in the formula, P i A first correlation coefficient between the ith monitoring class and the generated power, N is the total number of generated power vector elements, a n Is the nth element of the generated power vector, c in Is the nth element of the ith difference-like vector, b in Is the nth element of the ith monitoring class.
In an exemplary embodiment of the present invention, a plurality of sample vectors are first obtained, each sample vector including the generated power of the distributed power generation equipment and a plurality of monitoring data corresponding to a plurality of monitoring classes. For example, for a specific photovoltaic panel, the generated power at the first node time is acquired, and a plurality of monitoring data, such as the solar radiation intensity, the air temperature, the wind speed, and the like, at the first node time are also acquired. The plurality of monitoring data correspond to a plurality of monitoring classes, and the sample vectors are obtained by sequencing according to a preset sequence.
And a plurality of sample vectors are usually obtained for the same photovoltaic power generation equipment based on the sequence of time nodes.
As known, the photovoltaic power generation equipment is influenced by data of some monitoring classes, and whether other monitoring data influence the power generation power of the photovoltaic power generation equipment is seemingly uncertain, so that according to the embodiment of the invention, the monitoring vectors corresponding to a plurality of monitoring classes are constructed by sequentially obtaining the monitoring classes according to the time sequence in the sample vector, and similarly, the power generation power is arranged according to the time sequence to obtain the power generation power vector, and according to the relevance between the power generation power vector and the power generation power vector, the factor with low relevance can be removed.
An association determination method is an association coefficient formula, and an embodiment of the present invention provides a determination formula:
in the formula, P i A first correlation coefficient between the ith monitoring class and the generated power, N is the total number of generated power vector elements, a n Is the n-th element of the generated power vector, c in Is the n-th element of the i-th difference-like vector, b in Is the nth element of the ith monitoring class.
When the correlation coefficient calculated by the monitoring class through the above formula is lower than the threshold, it indicates that the monitoring class has little influence on the generated power and should be removed.
In step 103, adjusting a plurality of parameters of the photovoltaic power generation initial model according to the plurality of sample vectors to obtain a photovoltaic power generation power prediction model, wherein the photovoltaic power generation initial model is constructed based on an artificial neural network.
In some embodiments, the initial model of photovoltaic power generation comprises: the monitoring system comprises an input layer, a hidden layer and an output node, wherein the number of nodes of the input layer is the same as that of a plurality of monitoring data in a sample vector, the hidden layer comprises a plurality of fully-connected intermediate nodes, the input of the hidden layer is connected with the input layer, and the output node is connected with the output of the hidden layer;
the nodes corresponding to the memory classes in the input layer are constructed based on an LSTM model, wherein the memory classes are monitoring classes influencing the generated power by long-term data,
the step 103 comprises: arranging the plurality of sample vectors according to the time sequence;
a sample input step: inputting a plurality of monitoring data of each sample vector in the plurality of sample vectors into the photovoltaic power generation initial model to obtain a plurality of model outputs;
obtaining residual errors of the model outputs according to the model outputs and the generated power data of the sample vectors;
if the residual error is larger than the threshold value, adjusting a plurality of parameters of the photovoltaic power generation initial model according to the residual error and a gradient descent algorithm, and skipping to the sample input step;
and if not, taking the photovoltaic power generation initial model as the photovoltaic power generation power prediction model.
In some embodiments, the memory class is determined according to the following steps:
arranging the plurality of sample vectors according to the time sequence;
respectively obtaining a power generation power vector and a plurality of monitoring class vectors according to the plurality of sample vectors, wherein a plurality of elements of the power generation power vector are obtained based on power generation data of the plurality of sample vectors, and a plurality of elements of the monitoring class vectors are obtained based on monitoring data of the same monitoring class in the plurality of sample vectors;
obtaining a plurality of accumulated vectors according to the plurality of monitoring class vectors and a second formula, wherein the second formula is as follows:
in the formula, ar i (n + j) is the n + j element of the ith monitoring class, b in Is the nth element of the ith monitoring class, and j is the accumulation order;
obtaining a plurality of second correlation coefficients according to the generated power vector and the accumulated vectors, wherein the second correlation coefficients correspond to the monitoring classes, and the second correlation coefficients represent the correlation between the long-term data of the monitoring classes and the generated power;
and taking the monitoring class with the absolute value of the second correlation coefficient higher than the threshold value as the memory class.
Illustratively, the sample vector subjected to data reduction trains an artificial neural network, so as to obtain a prediction model.
Firstly, an initial model of photovoltaic power generation needs to be constructed, as shown in fig. 2, the initial model includes: the monitoring system comprises an input layer 201, a hidden layer 202 and an output node 203, wherein the number of the nodes of the input layer 201 is the same as that of a plurality of monitoring classes, the hidden layer 202 comprises a plurality of fully connected intermediate nodes, the input of the hidden layer 202 is connected with the input layer 201, and the output node 203 is connected with the output of the hidden layer 202.
For example, in the embodiment of the present invention, as shown in fig. 2, for a monitoring class affected by Long-Term data, an input point 2011 of a model corresponding to the monitoring class affected by Long-Term data is constructed based on an LSTM (Long Short-Term Memory network) model.
Compared with the original recurrent neural network, the LSTM solves the problem of gradient disappearance of a common network, can process long-sequence data and becomes one of the most popular network models at present.
After the model is constructed, a plurality of sample vectors can be input into the photovoltaic power generation initial model, specifically, a plurality of monitoring data of each sample vector are input into the photovoltaic power generation initial model to obtain the output of the model, the output of the model and the power generation data in the sample vectors are calculated to obtain a residual error, when the residual error is larger than a threshold value, a plurality of parameters of the model are adjusted according to a gradient descent method, then the input is carried out again, the iteration is carried out in such a way until the residual error of the model is lower than the threshold value, the parameters of the model are fixed, and the model is used as a photovoltaic power generation power prediction model.
For the determination method of the memory class, one method is to accumulate the monitoring vectors of the same class arranged according to the time sequence, and specifically, the second formula may be referred to:
in the formula, ar i (n + j) is the n + j element of the ith monitoring class, b in Is the nth element of the ith monitoring class, and j is the accumulation order;
the accumulation order may be set as required, for example, in the aspect of collecting the haze, if the collection interval is short, for example, once per hour, the accumulation order should be increased. Whereas if the acquisition interval is long, e.g. 3 acquisitions per 24 hours, the accumulation order should be reduced.
The relationship between the generated power and the long-term data can be determined by using the vector obtained by the above accumulation through a correlation coefficient calculation method, and if the generated power is affected by the long-term data, the corresponding input node should be set as the node of the LSTM model.
In a second aspect, an embodiment of the present invention provides a photovoltaic power generation power prediction method, including:
obtaining a photovoltaic power generation power prediction model constructed by the photovoltaic power generation power prediction model construction method according to the first aspect or any possible implementation manner of the first aspect;
acquiring a plurality of current data and a plurality of historical data sets according to the plurality of monitoring classes, wherein the historical data sets comprise a plurality of historical data, and the historical data is the historical monitoring data corresponding to the monitoring class lacking the current monitoring data;
generating a plurality of gray data corresponding to the current moment according to the plurality of historical data sets;
and inputting the current data and the grey data into the photovoltaic power generation power prediction model to obtain the predicted power generation power.
In one possible implementation, the generating a plurality of gray data corresponding to a current time from the plurality of historical data sets includes:
for each historical data set, the following steps are performed:
obtaining a sum vector according to a plurality of historical data of a historical data set and a third formula, wherein the third formula is as follows:
in the formula, h 1 (k) Is the kth element of the sum vector, K is the total number of elements in the historical dataset, h 0 (k) Is the mth element of the historical data set;
obtaining a mean vector according to the sum vector and a fourth formula, wherein the fourth formula is as follows:
in the formula, av 1 (k) The kth element of the mean vector;
determining a first weight and a second weight according to the sum vector, the mean vector and a fifth formula, wherein the fifth formula is as follows:
in the formula, A 1 First weight, A 2 Is a second weight, h 0 (K) The Kth element of the historical data set;
determining gray data according to the first weight, the second weight, the historical data set, and a sixth formula, wherein the sixth formula is:
wherein e is a natural constant, h 0 (K + 1) is gray data.
Illustratively, the second aspect of the present invention also provides a method for constructing a model by using the method of the first aspect, and as is known, some data are not easy to obtain directly, for example, in some scenarios, humidity data can be roughly determined by referring to previous humidity data through the internet although humidity data cannot directly obtain current humidity, so as to complement the data.
And inputting the complemented data into the power generation prediction model to obtain more accurate power generation prediction data.
The embodiment of the invention provides a data complementing mode, and specifically, historical data is firstly acquired, a sum vector is acquired through a third formula:
in the formula, h 1 (k) Is in the direction ofThe kth element of the quantity, K being the total number of elements in the historical dataset, h 0 (k) The mth element of the historical data set.
And obtaining a mean value vector according to the sum vector and a fourth formula, wherein the fourth formula is as follows:
in the formula, av 1 (k) The k-th element of the mean vector
Finally, according to a fifth formula and a sixth formula, gray data serving as complementary data is obtained:
the fifth formula is:
in the formula, A 1 First weight, A 2 Is a second weight, h 0 (K) The Kth element of the historical data set;
the sixth formula is:
wherein e is a natural constant, h 0 (K + 1) is gray data.
The embodiment of the photovoltaic power generation power prediction model construction method comprises the steps of firstly obtaining a plurality of sample vectors, wherein the sample vectors comprise power generation power data of target photovoltaic power generation equipment and a plurality of monitoring data obtained based on a plurality of monitoring classes, and the monitoring data and the power generation power data are obtained based on the same time node; then, removing data corresponding to irrelevant monitoring classes in the plurality of sample vectors according to the relevance between the generated power and the plurality of monitoring classes, wherein the relevance between the irrelevant monitoring classes and the generated power is lower than a threshold value; and finally, adjusting a plurality of parameters of the photovoltaic power generation initial model according to the plurality of sample vectors to obtain a photovoltaic power generation power prediction model, wherein the photovoltaic power generation initial model is constructed on the basis of an artificial neural network. According to the embodiment of the invention, when the generated power is predicted, some data irrelevant to the generated power are removed according to the generated power, so that the complexity of the model is reduced, and the calculation amount of model construction is reduced. The embodiment of the invention also sets the input layer of the LSTM node according to the relevance of the generated power and the long-term data, and the influence of the long-term data is considered, so that the model is closer to the physical reality and is more accurate to construct, thereby ensuring the prediction precision of the model.
In the aspect of photovoltaic power generation power prediction, firstly, a plurality of current data and a plurality of historical data sets are obtained according to a plurality of monitoring classes, wherein the historical data sets comprise a plurality of historical data, and the historical data is the historical monitoring data corresponding to the monitoring class lacking the current monitoring data; then, generating a plurality of gray data corresponding to the current moment according to the plurality of historical data sets; and finally, inputting the current data and the grey data into the photovoltaic power generation power prediction model to obtain the predicted power generation power. When the prediction model is used, some data which cannot be directly obtained are complemented, so that the number of the construction of monitoring points can be reduced, the consumption of resources is reduced, and the accuracy of the predicted data is ensured.
It should be understood that the sequence numbers of the steps in the above embodiments do not mean the execution sequence, and the execution sequence of each process should be determined by the function and the inherent logic, and should not constitute any limitation to the implementation process of the embodiments of the present invention.
The following are apparatus embodiments of the invention, and for details not described in detail therein, reference may be made to the corresponding method embodiments described above.
Fig. 3 is a functional block diagram of a photovoltaic power generation power prediction model building device according to an embodiment of the present invention, and referring to fig. 3, the photovoltaic power generation power prediction model building device 3 includes: a sample vector acquisition module 301, a data removal module 302, and a model construction module 303, wherein:
a sample vector obtaining module 301, configured to obtain a plurality of sample vectors, where a sample vector includes generated power data of a target photovoltaic power generation device and a plurality of monitoring data obtained based on a plurality of monitoring classes, and the plurality of monitoring data and the generated power data are obtained based on the same time node;
a data removing module 302, configured to remove data corresponding to an unrelated monitoring class from the plurality of sample vectors according to a correlation between the generated power and the plurality of monitoring classes, where the correlation between the unrelated monitoring class and the generated power is lower than a threshold;
the model building module 303 is configured to adjust a plurality of parameters of the photovoltaic power generation initial model according to the plurality of sample vectors to obtain a photovoltaic power generation power prediction model, where the photovoltaic power generation initial model is built based on an artificial neural network.
Fig. 4 is a functional block diagram of a terminal according to an embodiment of the present invention. As shown in fig. 4, the terminal 4 of this embodiment includes: a processor 400 and a memory 401, said memory 401 having stored therein a computer program 402 executable on said processor 400. The processor 400 executes the computer program 402 to implement the above-mentioned photovoltaic power generation prediction model construction method and embodiment, and the steps in the photovoltaic power generation prediction method and embodiment, such as steps 101 to 103 shown in fig. 1.
Illustratively, the computer program 402 may be partitioned into one or more modules/units, which are stored in the memory 401 and executed by the processor 400 to implement the present invention.
The terminal 4 may be a computing device such as a desktop computer, a notebook, a palm computer, and a cloud server. The terminal 4 may include, but is not limited to, a processor 400, a memory 401. Those skilled in the art will appreciate that fig. 4 is merely an example of a terminal 4 and does not constitute a limitation of terminal 4, and may include more or fewer components than those shown, or some components may be combined, or different components, e.g., terminal 4 may also include input-output devices, network access devices, buses, etc.
The Processor 400 may be a Central Processing Unit (CPU), other general purpose Processor, a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field Programmable Gate Array (FPGA) or other Programmable logic device, discrete Gate or transistor logic, discrete hardware components, etc. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.
The memory 401 may be an internal storage unit of the terminal 4, such as a hard disk or a memory of the terminal 4. The memory 401 may also be an external storage device of the terminal 4, such as a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) Card, a Flash memory Card (Flash Card), and the like, which are provided on the terminal 4. Further, the memory 401 may also include both an internal storage unit and an external storage device of the terminal 4. The memory 401 is used for storing the computer program 402 and other programs and data required by the terminal 4. The memory 401 may also be used to temporarily store data that has been output or is to be output.
It should be clear to those skilled in the art that, for convenience and simplicity of description, the foregoing division of the functional units and modules is only used for illustration, and in practical applications, the above function distribution may be performed by different functional units and modules as needed, that is, the internal structure of the apparatus may be divided into different functional units or modules to perform all or part of the above described functions. Each functional unit and module in the embodiments may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit, and the integrated unit may be implemented in a form of hardware, or may be implemented in a form of software functional unit. In addition, specific names of the functional units and modules are only for convenience of distinguishing from each other, and are not used for limiting the protection scope of the present application. The specific working processes of the units and modules in the system may refer to the corresponding processes in the foregoing method embodiments, and are not described herein again.
In the above embodiments, the description of each embodiment is focused on, and for parts that are not described or illustrated in detail in a certain embodiment, reference may be made to the description of other embodiments.
Those of ordinary skill in the art will appreciate that the various illustrative elements and algorithm steps described in connection with the embodiments disclosed herein may be implemented as electronic hardware or combinations of computer software and electronic hardware. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the implementation. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present invention.
In the embodiments provided in the present invention, it should be understood that the disclosed apparatus/terminal and method may be implemented in other manners. For example, the above-described apparatus/terminal embodiments are merely illustrative, and for example, the division of the modules or units is only one logical division, and there may be other divisions in actual implementation, for example, multiple units or components may be combined or integrated into another system, or some features may be omitted, or not executed. In addition, the shown or discussed mutual coupling or direct coupling or communication connection may be an indirect coupling or communication connection through some interfaces, devices or units, and may be in an electrical, mechanical or other form.
The units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the embodiment.
In addition, functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit. The integrated unit can be realized in a form of hardware, and can also be realized in a form of a software functional unit.
The integrated modules/units, if implemented in the form of software functional units and sold or used as separate products, may be stored in a computer readable storage medium. Based on such understanding, all or part of the flow in the method according to the above embodiments may be implemented by a computer program, which may be stored in a computer readable storage medium and used by a processor to implement the steps of the method and apparatus embodiments. Wherein the computer program comprises computer program code, which may be in the form of source code, object code, an executable file or some intermediate form, etc. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, usb disk, removable hard disk, magnetic disk, optical disk, computer Memory, read-Only Memory (ROM), random Access Memory (RAM), electrical carrier wave signals, telecommunications signals, software distribution medium, and the like.
The above embodiments are only used for illustrating the technical solutions of the present invention, and not for limiting the same; although the present invention has been described in detail with reference to the foregoing embodiments, it should be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may be modified or some technical features may be equivalently replaced; such modifications and substitutions do not substantially depart from the spirit and scope of the embodiments of the present invention, and are intended to be included within the scope of the present invention.
Claims (10)
1. A photovoltaic power generation power prediction model construction method is characterized by comprising the following steps:
obtaining a plurality of sample vectors, wherein each sample vector comprises generated power data of a target photovoltaic power generation device and a plurality of monitoring data obtained based on a plurality of monitoring classes, and the plurality of monitoring data and the generated power data are obtained based on the same time node;
removing data corresponding to irrelevant monitoring classes in the plurality of sample vectors according to the relevance between the generated power and the plurality of monitoring classes, wherein the relevance between the irrelevant monitoring classes and the generated power is lower than a threshold value;
and adjusting a plurality of parameters of the photovoltaic power generation initial model according to the plurality of sample vectors to obtain a photovoltaic power generation power prediction model, wherein the photovoltaic power generation initial model is constructed based on an artificial neural network.
2. The method for building the photovoltaic power generation power prediction model according to claim 1, wherein the removing data corresponding to irrelevant monitoring classes in the plurality of sample vectors according to the relevance between the power generation power and the plurality of monitoring classes comprises:
sorting the plurality of sample vectors in a predetermined order;
according to the plurality of sample vectors, respectively obtaining a power generation power vector and a plurality of monitoring class vectors, wherein a plurality of elements of the power generation power vector are obtained based on power generation data of the plurality of sample vectors, and a plurality of elements of the monitoring class vectors are obtained based on monitoring data of the same monitoring class in the plurality of sample vectors;
obtaining a plurality of first correlation coefficients according to the generated power vector and the plurality of monitoring class vectors, wherein the plurality of first correlation coefficients correspond to the plurality of monitoring classes, and the first correlation coefficients represent the correlation between the monitoring classes and the generated power;
taking the monitoring class with the absolute value of the first correlation coefficient lower than the threshold value as the irrelevant monitoring class;
and removing the monitoring data of the corresponding irrelevant monitoring class in the plurality of sample vectors.
3. The method for building the photovoltaic power generation power prediction model according to claim 2, wherein the obtaining a plurality of first correlation coefficients according to the power generation power vector and the plurality of monitoring class vectors comprises:
obtaining the plurality of first correlation coefficients according to a first formula, the generated power vector and the plurality of monitoring class vectors, wherein the first formula is as follows:
in the formula, P i A first correlation coefficient between the ith monitoring class and the generated power, N is the total number of generated power vector elements, a n Is the n-th element of the generated power vector, c in Is the n-th element of the i-th difference-like vector, b in Is the nth element of the ith monitoring class.
4. The photovoltaic power generation power prediction model construction method according to any one of claims 1 to 3, wherein the photovoltaic power generation initial model comprises: the monitoring system comprises an input layer, a hidden layer and an output node, wherein the number of nodes of the input layer is the same as that of a plurality of monitoring data in a sample vector, the hidden layer comprises a plurality of fully-connected intermediate nodes, the input of the hidden layer is connected with the input layer, and the output node is connected with the output of the hidden layer;
the nodes corresponding to the memory classes in the input layer are constructed based on an LSTM model, wherein the memory classes are monitoring classes of long-term data influencing the generated power;
adjusting a plurality of parameters of the photovoltaic power generation initial model according to the plurality of sample vectors to obtain a photovoltaic power generation power prediction model, including:
arranging the plurality of sample vectors according to the time sequence;
a sample input step: inputting a plurality of monitoring data of each sample vector in the plurality of sample vectors into the photovoltaic power generation initial model to obtain a plurality of model outputs;
obtaining residual errors of the model outputs according to the model outputs and the generated power data of the sample vectors;
if the residual error is larger than the threshold value, adjusting a plurality of parameters of the initial photovoltaic power generation model according to the residual error and a gradient descent algorithm, and skipping to the sample input step;
and if not, taking the photovoltaic power generation initial model as the photovoltaic power generation power prediction model.
5. The photovoltaic power generation power prediction model construction method according to claim 4, wherein the memory class is determined according to the following steps:
arranging the plurality of sample vectors according to the time sequence;
according to the plurality of sample vectors, respectively obtaining a power generation power vector and a plurality of monitoring class vectors, wherein a plurality of elements of the power generation power vector are obtained based on power generation data of the plurality of sample vectors, and a plurality of elements of the monitoring class vectors are obtained based on monitoring data of the same monitoring class in the plurality of sample vectors;
obtaining a plurality of accumulated vectors according to the plurality of monitoring class vectors and a second formula, wherein the second formula is as follows:
in the formula, ar i (n + j) is the n + j element of the ith monitoring class, b in Is the nth element of the ith monitoring class, and j is the accumulation order;
obtaining a plurality of second correlation coefficients according to the generated power vector and the accumulated vectors, wherein the second correlation coefficients correspond to the monitoring classes, and the second correlation coefficients represent the correlation between the long-term data of the monitoring classes and the generated power;
and taking the monitoring class with the absolute value of the second correlation coefficient higher than the threshold value as the memory class.
6. A photovoltaic power generation power prediction method is characterized by comprising the following steps:
obtaining a photovoltaic power generation power prediction model constructed according to the photovoltaic power generation power prediction model construction method of any one of claims 1 to 5;
acquiring a plurality of current data and a plurality of historical data sets according to the plurality of monitoring classes, wherein the historical data sets comprise a plurality of historical data, and the historical data is the historical monitoring data corresponding to the monitoring class lacking the current monitoring data;
generating a plurality of gray data corresponding to the current moment according to the plurality of historical data sets;
and inputting the current data and the grey data into the photovoltaic power generation power prediction model to obtain the predicted power generation power.
7. The photovoltaic generated power prediction method of claim 6, wherein the generating a plurality of gray data corresponding to a current time from the plurality of historical data sets comprises:
for each historical data set, the following steps are performed:
obtaining a sum vector according to a plurality of historical data of a historical data set and a third formula, wherein the third formula is as follows:
in the formula, h 1 (k) Is the kth element of the sum vector, K is the total number of elements in the historical dataset, h 0 (k) Is the mth element of the historical data set;
obtaining a mean vector according to the sum vector and a fourth formula, wherein the fourth formula is as follows:
in the formula, av 1 (k) The kth element of the mean vector;
determining a first weight and a second weight according to the sum vector, the mean vector and a fifth formula, wherein the fifth formula is as follows:
in the formula, A 1 First weight, A 2 Is a second weight, h 0 (K) The Kth element of the historical data set;
determining gray data according to the first weight, the second weight, the historical data set, and a sixth formula, wherein the sixth formula is:
wherein, is a natural constant, h 0 (K + 1) is gray data.
8. A photovoltaic power generation power prediction model construction device for realizing the photovoltaic power generation power prediction model construction method according to any one of claims 1 to 5, the photovoltaic power generation power prediction model construction device comprising:
the system comprises a sample vector acquisition module, a data processing module and a data processing module, wherein the sample vector acquisition module is used for acquiring a plurality of sample vectors, the sample vectors comprise generated power data of target photovoltaic power generation equipment and a plurality of monitoring data acquired based on a plurality of monitoring classes, and the plurality of monitoring data and the generated power data are acquired based on the same time node;
the data removing module is used for removing data corresponding to irrelevant monitoring classes in the sample vectors according to the relevance between the generated power and the monitoring classes, wherein the relevance between the irrelevant monitoring classes and the generated power is lower than a threshold value;
and the number of the first and second groups,
and the model building module is used for adjusting a plurality of parameters of the photovoltaic power generation initial model according to the plurality of sample vectors to obtain a photovoltaic power generation power prediction model, wherein the photovoltaic power generation initial model is built on the basis of an artificial neural network.
9. A terminal comprising a memory and a processor, the memory having stored therein a computer program operable on the processor, wherein the processor, when executing the computer program, performs the steps of the method according to any of claims 1 to 7.
10. A computer-readable storage medium, in which a computer program is stored which, when being executed by a processor, carries out the steps of the method according to any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211407351.5A CN115796338A (en) | 2022-11-10 | 2022-11-10 | Photovoltaic power generation power prediction model construction and photovoltaic power generation power prediction method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211407351.5A CN115796338A (en) | 2022-11-10 | 2022-11-10 | Photovoltaic power generation power prediction model construction and photovoltaic power generation power prediction method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115796338A true CN115796338A (en) | 2023-03-14 |
Family
ID=85436712
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211407351.5A Pending CN115796338A (en) | 2022-11-10 | 2022-11-10 | Photovoltaic power generation power prediction model construction and photovoltaic power generation power prediction method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115796338A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117313958A (en) * | 2023-11-28 | 2023-12-29 | 山东大学 | Low-voltage distributed photovoltaic power prediction method and system based on transfer learning |
-
2022
- 2022-11-10 CN CN202211407351.5A patent/CN115796338A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117313958A (en) * | 2023-11-28 | 2023-12-29 | 山东大学 | Low-voltage distributed photovoltaic power prediction method and system based on transfer learning |
| CN117313958B (en) * | 2023-11-28 | 2024-02-20 | 山东大学 | Low-voltage distributed photovoltaic power prediction method and system based on transfer learning |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110232483B (en) | Deep learning load prediction method and device and terminal equipment | |
| CN111027732B (en) | Method and system for generating multi-wind power plant output scene | |
| CN111079351B (en) | Power distribution network probability power flow acquisition method and device considering wind power uncertainty | |
| CN108063456B (en) | Distributed photovoltaic power generation grid-connected planning method and terminal equipment | |
| CN113553755B (en) | Power system state estimation method, device and equipment | |
| CN116707331B (en) | Inverter output voltage high-precision adjusting method and system based on model prediction | |
| CN111461445B (en) | Short-term wind speed prediction method and device, computer equipment and storage medium | |
| CN113516275A (en) | Power distribution network ultra-short term load prediction method and device and terminal equipment | |
| CN113936192A (en) | Power distribution network synchronous measurement missing data repairing method, terminal and storage medium | |
| CN114911788B (en) | Data interpolation method and device and storage medium | |
| CN115660233A (en) | Photovoltaic power prediction method and device, electronic equipment and storage medium | |
| CN115796338A (en) | Photovoltaic power generation power prediction model construction and photovoltaic power generation power prediction method | |
| CN112308335A (en) | Short-term electricity price prediction method and device based on xgboost algorithm | |
| CN117744916A (en) | Method and device for predicting energy storage capacity, computer equipment and readable storage medium | |
| CN110007371A (en) | Wind speed forecasting method and device | |
| CN116739172A (en) | Method and device for ultra-short-term prediction of offshore wind power based on climbing identification | |
| CN114971062A (en) | Photovoltaic power prediction method and device | |
| CN113779861B (en) | Photovoltaic Power Prediction Method and Terminal Equipment | |
| CN113190429B (en) | Server performance prediction method and device and terminal equipment | |
| CN111539573B (en) | Power prediction method and system for wind-solar hybrid off-grid system | |
| CN114971053A (en) | Training method and device for online prediction model of network line loss rate of low-voltage transformer area | |
| CN111612289A (en) | Electric power system risk assessment method based on new energy multi-scene risk feature guidance | |
| CN117996863B (en) | Sagging control method and device for photovoltaic power supply, computer equipment and medium | |
| CN116402160A (en) | Training method of power distribution network voltage prediction model and power distribution network voltage prediction method | |
| CN115936229A (en) | New energy output prediction method and device, electronic equipment and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |