CN116882597B - Virtual power plant control method, device, electronic equipment and readable medium - Google Patents
Virtual power plant control method, device, electronic equipment and readable medium Download PDFInfo
- Publication number
- CN116882597B CN116882597B CN202311147299.9A CN202311147299A CN116882597B CN 116882597 B CN116882597 B CN 116882597B CN 202311147299 A CN202311147299 A CN 202311147299A CN 116882597 B CN116882597 B CN 116882597B
- Authority
- CN
- China
- Prior art keywords
- load data
- matrix
- predicted
- data
- daily
- 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.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 46
- 239000011159 matrix material Substances 0.000 claims abstract description 211
- 238000012216 screening Methods 0.000 claims abstract description 28
- 230000004044 response Effects 0.000 claims description 29
- 230000008859 change Effects 0.000 claims description 24
- 238000012545 processing Methods 0.000 claims description 21
- 238000004590 computer program Methods 0.000 claims description 9
- 238000010248 power generation Methods 0.000 claims description 7
- 230000001174 ascending effect Effects 0.000 claims description 2
- 238000001914 filtration Methods 0.000 claims description 2
- 230000008569 process Effects 0.000 abstract description 10
- 239000002699 waste material Substances 0.000 abstract description 10
- 239000002245 particle Substances 0.000 description 47
- 238000005457 optimization Methods 0.000 description 11
- 230000006870 function Effects 0.000 description 9
- 238000010606 normalization Methods 0.000 description 7
- 238000004891 communication Methods 0.000 description 6
- 238000010586 diagram Methods 0.000 description 6
- 230000008901 benefit Effects 0.000 description 4
- 230000002159 abnormal effect Effects 0.000 description 3
- 230000005540 biological transmission Effects 0.000 description 3
- 238000004146 energy storage Methods 0.000 description 3
- 230000003287 optical effect Effects 0.000 description 3
- 238000013468 resource allocation Methods 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 2
- 238000005259 measurement Methods 0.000 description 2
- 230000000644 propagated effect Effects 0.000 description 2
- 230000000295 complement effect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 239000000835 fiber Substances 0.000 description 1
- 230000008014 freezing Effects 0.000 description 1
- 238000007710 freezing Methods 0.000 description 1
- 238000012804 iterative process Methods 0.000 description 1
- 239000004973 liquid crystal related substance Substances 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 239000013307 optical fiber Substances 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/06—Energy or water supply
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Economics (AREA)
- Human Resources & Organizations (AREA)
- Theoretical Computer Science (AREA)
- Strategic Management (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Marketing (AREA)
- Tourism & Hospitality (AREA)
- Data Mining & Analysis (AREA)
- General Business, Economics & Management (AREA)
- Entrepreneurship & Innovation (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Game Theory and Decision Science (AREA)
- Development Economics (AREA)
- Health & Medical Sciences (AREA)
- Educational Administration (AREA)
- Water Supply & Treatment (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Public Health (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The embodiment of the disclosure discloses a virtual power plant control method, a virtual power plant control device, electronic equipment and a readable medium. One embodiment of the method comprises the following steps: acquiring a historical daily load data set and predicted daily known period load data of a virtual power plant; according to weather information of the load data of the known period of the predicted day and the predicted day type, carrying out data screening on the historical daily load data subset to obtain a screened similar daily load data set; generating a load data matrix according to the screened similar daily load data set and the load data of the predicted daily known period; inputting the load data matrix into a load data prediction model to obtain load data of a prediction day prediction period; and carrying out cooperative scheduling control on the virtual power plant according to the load data of the forecast day forecast period. The embodiment reduces the waste of power resources in the scheduling process.
Description
Technical Field
Embodiments of the present disclosure relate to the field of computer technology, and in particular, to a virtual power plant control method, apparatus, electronic device, and readable medium.
Background
The virtual power plant is used as the final configuration of the energy Internet, and a green and intelligent development mode can be provided for the power system. Currently, the virtual power plant is controlled in the following manner: and (5) short-term predicting load data of the virtual power plant by using a time sequence method to realize control of the virtual power plant.
However, when the above manner is adopted, there are often the following technical problems:
first, because fluctuation law of load data change is difficult to predict, so that scheduling control resource allocation is unreasonable, when the scheduling control resource is too large, a power line with certain conveying capacity may not bear excessive electric energy, transmission loss is caused, and power resource waste is caused in the scheduling process.
Second, the virtual power plant cannot be differentially powered in different time periods, so that more power loss is generated when the virtual power plant is powered due to insufficient power supply or excessive power supply, and abnormal distribution may occur when the energy storage resource distributes power resources, so that the stability of the virtual power plant control is lower.
The above information disclosed in this background section is only for enhancement of understanding of the background of the inventive concept and, therefore, may contain information that does not form the prior art that is already known to those of ordinary skill in the art in this country.
Disclosure of Invention
The disclosure is in part intended to introduce concepts in a simplified form that are further described below in the detailed description. The disclosure is not intended to identify key features or essential features of the claimed subject matter, nor is it intended to be used to limit the scope of the claimed subject matter.
Some embodiments of the present disclosure propose virtual power plant control methods, apparatus, electronic devices, and computer readable media to address one or more of the technical problems mentioned in the background section above.
In a first aspect, some embodiments of the present disclosure provide a virtual power plant control method, the method comprising: acquiring a historical daily load data set and predicted daily known period load data of a virtual power plant; clustering the historical daily load data set and the predicted daily known period load data to obtain a clustering result; according to the clustering result, a historical daily load data subset which has a load change rule similarity relationship after clustering with the load data of the known period of the predicted daily load is screened from the historical daily load data set; according to the weather information and the predicted day type of the predicted day known period load data, carrying out data screening on the historical day load data subset to obtain a screened similar day load data set; generating a load data matrix according to the screened similar daily load data set and the load data of the predicted daily known period; inputting the load data matrix into a load data prediction model to obtain load data of a prediction day prediction period; and carrying out cooperative scheduling control on the virtual power plant according to the forecast day forecast period load data.
In a second aspect, some embodiments of the present disclosure provide a virtual power plant control apparatus, the apparatus comprising: an acquisition unit configured to acquire a historical daily load data set and predicted daily known period load data of the virtual power plant; a clustering unit configured to cluster the historical daily load data set and the predicted daily known period load data to obtain a clustering result; a first screening unit configured to screen, from the historical daily load data set, a subset of historical daily load data having a similar relationship to the post-cluster load change rule between the load data of the predicted daily known period according to the clustering result; the second screening unit is configured to perform data screening on the historical daily load data subset according to weather information and the predicted daily type of the load data in the known period of the predicted day to obtain a screened similar daily load data set; a generation unit configured to generate a load data matrix based on the filtered similar daily load data set and the predicted daily known period load data; an input unit configured to input the load data matrix to a load data prediction model to obtain predicted day prediction period load data; and the control unit is configured to perform cooperative scheduling control on the virtual power plant according to the forecast day forecast period load data.
In a third aspect, some embodiments of the present disclosure provide an electronic device comprising: one or more processors; a storage device having one or more programs stored thereon, which when executed by one or more processors causes the one or more processors to implement the method described in any of the implementations of the first aspect above.
In a fourth aspect, some embodiments of the present disclosure provide a computer readable medium having a computer program stored thereon, wherein the program, when executed by a processor, implements the method described in any of the implementations of the first aspect above.
The above embodiments of the present disclosure have the following advantages: by the virtual power plant control method of some embodiments of the present disclosure, waste of power resources in the scheduling process is reduced. Specifically, the reason why the waste of the power resource is caused is that: because fluctuation law of load data change is difficult to predict, scheduling control resource allocation is unreasonable, when the scheduling control resource is overlarge, a power line with certain conveying capacity can not bear excessive electric energy, transmission loss is caused, and power resource waste is caused in the scheduling process of the power resource. Based on this, the virtual power plant control method of some embodiments of the present disclosure first obtains a historical daily load dataset and a predicted daily known period load data of the virtual power plant. Thus, a historical daily load data set of the virtual power plant and a predicted daily known period load data can be obtained for subsequent operation. And then, clustering the historical daily load data set and the predicted daily known period load data to obtain a clustering result. Therefore, the two data with higher similarity can be continuously combined until the clustering result is obtained after the combination. And then, according to the clustering result, a historical daily load data subset which has a similar relation with the load change rule after clustering between the load data in the known period of the predicted daily load and the historical daily load data set is screened. Thus, a subset of the historical daily load data that is the same class as the predicted daily known period load data can be selected from the clustering result so that the range similar to the predicted daily known period load data is narrowed. And then, according to the weather information and the predicted day type of the predicted day known period load data, carrying out data screening on the historical day load data subset to obtain a screened similar day load data set. Therefore, screening can be further carried out according to weather information and the predicted day type, so that a screened similar day load data set is obtained, the prediction error in predicting the load data of the virtual power plant is reduced, and then a load data matrix is generated according to the screened similar day load data set and the load data of the predicted day known period. Thus, the filtered similar daily load data set and the predicted daily known period load data can be formed into a load data matrix. And then, inputting the load data matrix into a load data prediction model to obtain the load data of the prediction day prediction period. Therefore, the load data matrix can be predicted according to the load data prediction model, and the data which is not filled in the load data matrix can be filled, so that the load data of the prediction day prediction period can be obtained. And finally, carrying out cooperative scheduling control on the virtual power plant according to the forecast day forecast period load data. Therefore, the waste of the power resources in the scheduling process is reduced.
Drawings
The above and other features, advantages, and aspects of embodiments of the present disclosure will become more apparent by reference to the following detailed description when taken in conjunction with the accompanying drawings. The same or similar reference numbers will be used throughout the drawings to refer to the same or like elements. It should be understood that the figures are schematic and that elements and components are not necessarily drawn to scale.
FIG. 1 is a flow chart of some embodiments of a virtual power plant control method according to the present disclosure;
FIG. 2 is a schematic structural view of some embodiments of a virtual power plant control device according to the present disclosure;
fig. 3 is a schematic structural diagram of an electronic device suitable for use in implementing some embodiments of the present disclosure.
Detailed Description
Embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. While certain embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be embodied in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete. It should be understood that the drawings and embodiments of the present disclosure are for illustration purposes only and are not intended to limit the scope of the present disclosure.
It should be noted that, for convenience of description, only the portions related to the present invention are shown in the drawings. Embodiments of the present disclosure and features of embodiments may be combined with each other without conflict.
It should be noted that the terms "first," "second," and the like in this disclosure are merely used to distinguish between different devices, modules, or units and are not used to define an order or interdependence of functions performed by the devices, modules, or units.
It should be noted that references to "one", "a plurality" and "a plurality" in this disclosure are intended to be illustrative rather than limiting, and those of ordinary skill in the art will appreciate that "one or more" is intended to be understood as "one or more" unless the context clearly indicates otherwise.
The names of messages or information interacted between the various devices in the embodiments of the present disclosure are for illustrative purposes only and are not intended to limit the scope of such messages or information.
The present disclosure will be described in detail below with reference to the accompanying drawings in conjunction with embodiments.
FIG. 1 is a flow 100 of some embodiments of the virtual power plant control method of the present disclosure. The virtual power plant control method comprises the following steps:
Step 101, acquiring a historical daily load data set and predicted daily known period load data of a virtual power plant.
In some embodiments, an executing entity (e.g., a computing device) of the virtual power plant control method may obtain the historical daily load data set and the predicted daily known period load data of the virtual power plant by way of a wired connection or a wireless connection.
Here, the above-described virtual power plant may refer to a distributed resource and load coordination management system that participates in the power market and grid operation. The historical daily load data in the historical daily load data set may refer to load data generated daily in the past. The load data may refer to electric power data consumed by electric devices using electric energy or power data of all electric devices in an area at a certain moment. The above-described predicted day-known period load data may refer to load data of a known period of the predicted day. For example, the predicted day may be 2023, 8, 6, and the known time period of the predicted day may be 2023, 8, 6, 8 am: 00 to 11 am: 00.
it should be noted that the wireless connection may include, but is not limited to, 3G/4G connections, wiFi connections, bluetooth connections, wiMAX connections, zigbee connections, UWB (ultra wideband) connections, and other now known or later developed wireless connection means.
And 102, clustering the historical daily load data set and the predicted daily known period load data to obtain a clustering result.
In some embodiments, the executing entity may cluster the historical daily load data set and the predicted daily known period load data to obtain a clustered result.
Here, the clustering result may refer to a tree-structured clustering tree, where each leaf node of the clustering tree represents one data point, the internal node represents a cluster having a similarity relationship with a variation rule of load data, and the root node represents one cluster of all data points.
As an example, the execution body may first combine the historical daily load data set and the load data of the predicted daily known period, and then gradually combine each load data in the combined load data set as a load data class from bottom to top until the load data classes with similar change rules are combined into a cluster, so as to obtain a clustering result.
Optionally, the executing body may cluster the historical daily load data set and the predicted daily known period load data to obtain a clustering result by:
First, determining the load data of the known period of the predicted day and the load data set of the historic day as normalized load data sets.
As an example, the execution subject may first perform normalization processing on the load data of the known period of predicted day to obtain load data of the known period of predicted day after normalization, then perform normalization processing on the load data set of the historical day to obtain a load data set of the historical day after normalization, and finally combine the load data of the known period of predicted day after normalization with the load data set of the historical day after normalization to obtain a load data set of the historical day after normalization.
Second, the following clustering steps are performed on the normalized post-load dataset:
the first substep, determining the similarity between every two normalized post-load data in the normalized post-load data set, and obtaining a similarity set.
Here, the above-mentioned similarity may refer to euclidean distance.
And a second sub-step of determining the first normalized post-load data and the second normalized post-load data corresponding to the maximum similarity in the similarity set as normalized post-load data classes.
And a third sub-step of deleting the first normalized post-load data and the second normalized post-load data from the normalized post-load data set to obtain a deleted post-load data set.
And a fourth sub-step of determining each deleted load data in the deleted load data set as a deleted load data class to obtain a deleted load data class set.
And a fifth sub-step of combining the standardized post-load data class and the deleted post-load data class set into a clustered post-load data class set.
Here, the above-described combination may refer to a combination.
And a sixth substep of determining the post-cluster load data set as a normalized post-cluster load data set in response to determining that the number of post-cluster load data classes in the post-cluster load data set is equal to or greater than a preset number of clusters, and executing the clustering step again.
Here, the preset number of clusters may refer to 2.
And thirdly, determining the obtained clustered load data class set as a clustering result in response to the fact that the number of clustered load data classes in the clustered load data class set is smaller than the preset clustering number.
And 103, screening a historical daily load data subset which has a similar relationship with the load change rule after clustering between the load data in the known period of the predicted daily load from the historical daily load data set according to the clustering result.
In some embodiments, the execution entity may screen, according to the clustering result, a subset of the historical daily load data having a similar relationship to the post-cluster load change rule between the load data of the known period of time of the predicted day from the set of historical daily load data.
Here, the historical daily load data subset may refer to at least one load data having a similar relationship of post-clustering load change law with the load data of the above-described predicted daily known period among load data generated daily in the past. Here, the similarity relationship of the load change rule after clustering may refer to a similarity relationship that the shape similarity of the load change curve after clustering is greater than a preset threshold. For example, the preset threshold may refer to 0.8.
As an example, the execution subject may first compare the predicted daily known period load data with each of the historical daily load data in the set of historical daily load data, and then select at least one historical daily load data having a similar relationship with the post-cluster load change rule between the predicted daily known period load data from the clustering result, to obtain a subset of the historical daily load data.
And 104, carrying out data screening on the historical daily load data subset according to weather information and the predicted daily type of the load data in the known period of the predicted day to obtain a screened similar daily load data set.
In some embodiments, the execution subject may perform data screening on the subset of historical daily load data according to weather information and the type of the predicted daily load data in the known period of time of the predicted day, to obtain a screened similar daily load data set.
Here, the weather information may refer to various kinds of phenomenon information occurring in the atmosphere. For example, the weather information may include, but is not limited to, at least one of: sunny days, cloudy days and rainy days. The prediction day type may be a type to which a date corresponding to the prediction day belongs. For example, the above-described predicted day types may include, but are not limited to, at least one of: workday, holiday. Here, the post-screen similar daily load data in the above-described post-screen similar daily load data set may refer to similar daily load data having weather information and a predicted daily type similar to the above-described predicted daily known period load data.
As an example, the execution subject may first determine, as the post-screening similar daily load data set, each of the historical daily load data similar to the weather information and the predicted daily type, by comparing the weather information and the predicted daily type of the predicted daily load data with each of the historical daily load data in the subset of the historical daily load data.
Optionally, the executing body may perform data screening on the historical daily load data subset according to weather information and the predicted daily type of the load data in the known period of the predicted day, to obtain a screened similar daily load data set:
The first step is to determine the data similarity between each historical daily load data in the historical daily load data subset and the load data of the known period of the predicted daily according to the weather information and the predicted daily type, and obtain a data similarity set.
Here, the above data similarity may refer to euclidean distance.
And a second step of determining, for each data similarity in the data similarity set, historical daily load data corresponding to the data similarity greater than the preset data similarity as post-screening similar daily load data in response to determining that the data similarity is greater than the preset data similarity, and obtaining a post-screening similar daily load data set.
Here, the above-described preset data similarity may refer to 0.7.
And 105, generating a load data matrix according to the filtered similar daily load data set and the load data of the known period of the predicted day.
In some embodiments, the execution body may generate a load data matrix according to the filtered similar daily load data set and the predicted daily known period load data.
Here, the load data matrix may refer to a load data matrix including load data and zero elements. For example, the load data matrix may refer to 。/>Representing the normalized signature of the load data matrix. The load data matrix may be +.>. Wherein the above elements->Element->Element->Element->Element->Element->Element->Element->Element->The elements corresponding to the load data are respectively. The above elementsIs a zero element.
As an example, the execution subject may first normalize the filtered similar daily load data set and the predicted daily known period load data, then convert the normalized filtered similar daily load data set and the normalized predicted daily known period load data into a matrix, and then fill a region in the matrix where a portion to be predicted is filled with 0 to obtain a load data matrix.
And 106, inputting the load data matrix into a load data prediction model to obtain the load data of the prediction day prediction period.
In some embodiments, the execution body may input the load data matrix into a load data prediction model to obtain the load data of the prediction date prediction period.
Here, the load data prediction model may be referred to as GAIN (Generative Adversarial Imputation Network) model. The load data prediction model is a load data prediction model for data interpolation (interpolation). The load data prediction model aims to generate data which is close to the real load change rule as much as possible so as to fill the real data. The above-described predicted day predicted period load data may refer to load data of a predicted day predicted period. For example, the above predicted day prediction period load data may refer to 2023, 8, 6, and 13 pm: 00 to 17: load data of 00.
Optionally, the executing body may input the load data matrix into a load data prediction model to obtain the load data of the prediction day prediction period by the following steps:
and the first step is to input the load data matrix, the noise matrix and the mask matrix into a load data filling model in the load data prediction model to obtain a generating matrix.
Here, the above-mentioned noise matrix may mean that noise is added according to a portion of the load matrix that needs to be predicted. For example, the noise matrix may be. Wherein each element in the noise matrix may represent +.>. Above->Representing load data matrix->Line->Column elements. Above->Representing when the load data matrix is +.>Line->The element of the column is 0. I.e. when->When (I)>Equal to->. Above->The range of values of (2) is {1, n }. n is the number of daily load sample data points. Above->The range of values of (2) is {1, m }. The above m is the number of load days of the predicted day and its similar days. Above-mentionedCan be +.>Not equal to 0. The noise matrix may be +.>. Wherein, element->Element->Element->Element->Element->Element->The elements corresponding to the noise load data.
Here, the mask matrix may refer to a mask matrix that judges data. For example, the mask matrix may refer to . Wherein each element of the mask matrix may represent +.>. Above->Representing when the load data matrix is +.>Line->The elements of the column are not equal to 0. The mask matrix may be +.>. Wherein, element->Representing that the corresponding position in the load data matrix is the measured load data, element +.>The corresponding position in the representative load data matrix is unmeasured load data.
Here, the generator matrix may be a generator matrix predicted by a load data padding model in the load data prediction model. For example, the generator matrix is expressed as. Above->Representing the generator matrix. Above->A generator representing a load data prediction model. Above->Representing a Hadamard product operation, i.e. multiplication of two matrix corresponding position elements. Above->May refer to the complementary matrix of the mask matrix.
And secondly, filling the load data matrix according to the generated matrix to obtain an interpolation matrix, wherein the load data corresponding to matrix elements in the interpolation matrix comprises real data and false data.
Here, the interpolation matrix may refer to. The interpolation matrix is expressed as. The real data may refer to actual load data. The dummy data may refer to generated load data 。
And thirdly, generating the load data of the prediction day prediction period according to the interpolation matrix.
Optionally, after the load data matrix is input into the load data prediction model by the execution subject to obtain the load data of the prediction day prediction period, the method further includes:
and optimizing the load data prediction model according to the load data of the prediction day prediction period and the objective function to obtain an optimized load data prediction model.
Optionally, the executing body may optimize the load data prediction model according to the load data and the objective function of the prediction day prediction period to obtain an optimized load data prediction model:
the method comprises the steps of firstly, defining a load data noise matrix for each load data sample in a load data sample set, wherein the load data matrix and a mask matrix are load data matrix sample sets, and obtaining a load data matrix sample set.
Here, each of the above-described load data samples in the load data sample set may refer to a sample corresponding to the load data. The load data noise matrix may be a load data noise matrix in which the position containing the measurement data as an element in the load data matrix is set to 0 and noise is added to the position to be predicted.
Secondly, defining a prompt matrix aiming at a mask matrix of each load data sample in the load data sample set to obtain a prompt matrix set.
Here, the prompt matrix may refer to a matrix that prompts the arbiter to determine whether the data is real measurement data. For example, the hint matrix may refer to. Where element 0.5 represents uncertainty as to whether the load data was measured or false data was generated. Element 1 represents measured load data. Element 0 represents unmeasured load data. The prompt matrix is generated according to the mask matrix, and can be setThe number of elements 1 or 0 in the matrix is shown as the duty ratio, and the positions where the elements 1 and 0 are arranged in the prompt matrix are consistent with the mask matrix.
Third, for each of the load data matrix sample groups in the set of load data matrix sample groups, performing the following first processing step:
a first sub-step of inputting the load data noise matrix, the load data matrix, and the mask matrix in the load data matrix sample group into a load data padding model in the load data prediction model to obtain a load data interpolation matrix.
Here, the load data padding model in the load data prediction model may be a load data padding model in which the load data noise matrix, the load data matrix, and the mask matrix are input, and the load data interpolation matrix is output. The load data padding model in the load data prediction model may refer to a generator. The generator may refer to a generation network (Generator Network). The load data padding model in the load data prediction model may be a load data padding model for padding missing load data. The load data interpolation matrix may be a load data interpolation matrix obtained by filling a portion to be predicted in the load data matrix with data generated by the generator.
And a second substep, inputting the load data interpolation matrix and the prompt matrix in the prompt matrix set into a load data probability judging model in a load data prediction model to obtain a predicted mask matrix, wherein each element in the predicted mask matrix represents the probability of whether the load data is real data or not, and the prompt matrix is a prompt matrix of the load data samples in the prompt matrix set, which is the same as the load data mask matrix.
Here, the load data probability discrimination model in the above-described load data prediction model may be referred to as a discriminator. The arbiter may refer to a discrimination network (Discriminator Network). The load data probability discrimination model in the load data prediction model may be a load data probability discrimination model having a load data interpolation matrix and a presentation matrix as inputs and a predicted mask matrix as an output. The load data probability judging model in the load data predicting model can be used for judging the probability that the load data is real data in the load data noise matrix sample predicting information. The predicted mask matrix may refer to a probability that the load data corresponding to each element in the load data interpolation matrix output by the arbiter is real data.
And a third sub-step of adjusting network parameters of the load data padding model in response to determining that the load data interpolation matrix is not equal to the load data matrix.
Here, the adjustment may be to calculate a difference between the load data interpolation matrix and the load data matrix. On this basis, the difference value is transmitted forward from the last layer of the model by using back propagation, random gradient descent and the like to adjust the parameters of each layer. Of course, a network freezing (dropout) method may be used as needed, and network parameters of some layers therein may be kept unchanged and not adjusted, which is not limited in any way.
And a fourth sub-step of adjusting network parameters of the load data probability discrimination model in response to determining that the predicted mask matrix is not equal to the mask matrix.
And a fourth step of determining the load data filling model as a final load data filling model in response to determining that the difference between the data in the load data interpolation matrix and the data corresponding to the load data matrix is smaller than a preset first threshold.
Here, the preset first threshold may be a preset first threshold for the load data padding model. For example, the preset first threshold may refer to 0.002.
And fifthly, determining the load data probability discrimination model as a final load data probability discrimination model in response to determining that the difference between the predicted mask matrix and the mask matrix is smaller than a preset second threshold.
Here, the preset second threshold may refer to a preset second threshold for the load data probability discrimination model. The preset second threshold may refer to 0.01.
And sixthly, determining the final load data filling model and the final load data probability judging model as optimized load data prediction models.
And 107, performing cooperative scheduling control on the virtual power plant according to the predicted daily predicted period load data.
In some embodiments, the executing entity may perform cooperative scheduling control on the virtual power plant according to the predicted daily predicted period load data.
As an example, the executing body may first obtain an actual load peak shaving demand according to the load data of the predicted time period of the predicted day, and then adjust the generation of new energy according to the actual load peak shaving demand, so as to perform cooperative scheduling control on the virtual power plant.
Optionally, the executing body may perform cooperative scheduling control on the virtual power plant according to the predicted day and predicted period load data by:
and in the first step, in response to determining that a load change curve corresponding to the load data of the forecast day forecast period is an ascending curve, increasing the power generation amount of the virtual power plant unit so as to perform cooperative scheduling control on the virtual power plant.
And secondly, in response to determining that a load change curve corresponding to the load data of the forecast day forecast period is a descending curve, reducing the power generation amount of the virtual power plant unit so as to perform cooperative scheduling control on the virtual power plant.
Optionally, after the step 107, the method further includes:
and determining a load data predicted value of the load data of the predicted day prediction period as load data predicted information at intervals of preset time to obtain a load data predicted information set.
Here, the above-described every predetermined time may refer to a predetermined interval time. For example, the above every predetermined time may refer to 15 minutes. The load data prediction information in the load data prediction information set may represent the load data prediction value at the interval of the preset time. The load data prediction value may be a value for predicting the power load.
And a second step of generating a load data prediction curve based on the load data prediction information set.
As an example, the execution body may connect the load data prediction information in the load data prediction information set in time sequence to obtain the load data prediction curve.
And thirdly, setting an optimization target for the load data prediction curve, wherein the optimization target represents a unit combined output curve and the load data prediction curve are kept as consistent as possible, and the unit combined output curve represents a superposition curve of unit output curves of all units of the virtual power plant.
Here, the above-mentioned optimization target represents. Wherein (1)>Indicate->Every preset time. />Is in the range of [1, 96 ]]。/>Representing the optimization objective as distance +.>Is a minimum of (2). />Represent the firstPredicted values of the load data at each moment. />Indicate->And combining the output power generation data by the unit at each moment. />Representing from->Summation from 1 to 96.
As an example, the execution subject may set a mean square error (Mean Square Error, MSE) between the load data prediction curve and a unit combination output curve corresponding to unit power generation data of each unit of the virtual power plant as the optimization target. Wherein, the mean square error is smaller than a preset value. For example, the mean square error may refer to 0.001. The above-mentioned preset value may refer to 0.002.
And fourthly, determining the initialized particle sets according to each unit in the virtual power plant.
Here, the initialized particle set may refer to all possible scheduling schemes of each unit after determining initial output information of each unit. The scheduling scheme can refer to a distribution scheme of the output of each unit of the virtual power plant. For example, the above-described assembly may be referred to as a generator assembly.
As an example, the executing body may randomly generate initial output information of each unit according to an output range of each unit in the virtual power plant, and then obtain an initialized particle set according to a load demand and the output range of each unit. The output range may be a range between a minimum output range and a maximum output range, and the initialized particle set may be all possible unit scheduling schemes meeting the load requirement after determining initial output information of each unit. The random generation may refer to randomly generating a set of initial force values from the force range that meet the load demand. And in response to determining that the initial output value is smaller than the output upper limit value of the unit, adjusting the initialized output information of each unit according to the system load requirement to obtain an adjusted unit scheduling scheme, wherein the adjusted unit scheduling scheme is used as an initialized particle, and all initialized feasible scheduling schemes form an initialized particle set.
Fifth, for each particle in the initialized particle set, the following second processing step is performed:
a first sub-step of performing the following iterative processing steps on the particles:
step one, updating the initialized position information of the particles according to the initial position information, particle speed, historical optimal position information and global optimal position information of the particles.
Here, the above-mentioned particles may refer to a scheduling scheme of units in the virtual power plant. The initial particle position information may be an initial unit output information determined randomly. The particle velocity may refer to the change in the force exerted by the particle from the current position information to the next position information. The above-mentioned history optimal position information may refer to a history local optimal scheduling scheme. For example, the above-mentioned historical optimal location information may refer to a local optimal scheduling scheme for one unit in the virtual power plant that satisfies the part-time load demand found in past scheduling schemes. The global optimal position information may refer to a global optimal scheduling scheme capable of meeting load requirements. For example, the global optimal position information may refer to an optimal scheduling scheme found from all unit scheduling schemes of the virtual power plant under the condition that an optimization target that the unit combined output is consistent with the load demand is satisfied. The above-mentioned position information of the particles may refer to the current position information of the particles in each iteration.
As an example, the execution subject may first determine the velocity from initial position information of the particle, history optimal position information, and global optimal position information by inertial weight×particle velocity+learning factor 1×random number 1× (history optimal position information-particle initial position information) +learning factor 2×random number 2× (global optimal position information-particle initial position information). Wherein the inertia weight, the learning factor 1 and the learning factor 2 are experience parameters, and the random number 1 and the random number 2 are random numbers generated in a certain range. For example, the initial particle position information of the particles is 5, the history optimal position information is 8, and the global optimal position information is 10. The inertia weight is set to 0.2, the learning factor 1 is 0.5, and the learning factor 2 is 0.9. Random number 1 is 0.3, random number 2 is 0.7, and the particle velocity is 2 in the last iteration, then the velocity is 0.2x2+0.5x0.3× (8-5) +0.9x0.7× (10-5) =4. Then, the sum of the initial position information of the particles and the velocity is determined as updated position information of the particles. For example, the velocity is 4, the current position information is 5, i.e., 5+4=9, and the updated position information of the particles is 9.
And a second sub-step of determining the fitness of the particles according to the optimization target.
Here, the fitness may be a shape similarity between a unit combined output curve in the virtual power plant and a curve of the predicted daily predicted period load data.
As an example, the execution body may first use the unit output power generation data corresponding to the initial position of the particle as the optimization targetThen, the obtained value is taken as a fitness value.
And a third sub-step of continuing to execute the iterative processing step by updating the particles in response to determining that the number of iterative processing steps does not reach the preset iterative processing number.
Here, the above-described preset number of iterative processes may refer to 10 times.
And a second sub-step of determining the updated particles obtained after the preset iteration times as an optimal power scheduling scheme in response to determining that the number of the iteration processing steps reaches the preset iteration times.
And a third sub-step, updating the optimal power scheduling scheme according to the fitness of the particles obtained after the preset iteration times, and obtaining the updated optimal power scheduling scheme.
As an example, the execution body may update the location information of the particle corresponding to the obtained fitness value of the particle to the updated optimal power scheduling scheme by the obtained fitness value of the particle being greater than a preset fitness threshold.
And sixthly, superposing the set output curves corresponding to the updated optimal power dispatching scheme to obtain a superposed set output curve serving as a set combined output curve.
Here, the above-described superposition may mean that the unit output values of the same period of time are added.
And seventhly, adjusting the power storage resources according to the updated optimal power scheduling schemes to realize collaborative scheduling of the virtual power plants in response to the fact that the shape similarity of the combined output curve of the unit and the curve of the load data of the forecast day forecast period is higher than a preset similarity threshold.
Here, the adjustment may refer to charge/discharge adjustment of the power storage device.
The related matters in the first step to the ninth step are taken as an invention point of the disclosure, and the second technical problem mentioned in the background art is solved, namely the virtual power plant cannot be differentially powered in different time periods, so that more power loss is generated by insufficient power supply or excessive power supply when the virtual power plant is powered, and abnormal distribution may occur when the energy storage resource distributes the power resource, and the stability of the control of the virtual power plant is lower. The factors that cause the virtual power plant to generate more power loss due to insufficient power supply or excessive power supply when the virtual power plant is powered on and the stability of the virtual power plant control is lower are often as follows: the virtual power plants cannot be differentially powered in different time periods, so that more power loss is generated when the virtual power plants are powered due to insufficient power supply or excessive power supply, and abnormal distribution can occur when the energy storage resources are used for distributing the power resources, so that the stability of the virtual power plant control is lower. If the factors are solved, the effect of reducing the power loss generated by insufficient power supply or excessive power supply of the virtual power plant when the virtual power plant is powered can be achieved, and the stability of the control of the virtual power plant is improved. To achieve this, first, a load data prediction value at every predetermined time in the load data of the predicted day prediction period is determined as load data prediction information, and a load data prediction information set is obtained. Then, a load data prediction curve is generated based on the load data prediction information set. And setting an optimization target for the load data prediction curve, wherein the optimization target represents a unit combined output curve which represents a superposition curve of unit output curves of all units of the virtual power plant, and the load data prediction curve is kept as consistent as possible. And then, according to each unit in the virtual power plant, determining the initialized particle set. Thereafter, for each particle in the initialized particle set, the following second processing step is performed: the following iterative processing steps are performed on the particles: and updating the position information of the particles according to the initial position information, the particle speed, the historical optimal position information and the global optimal position information of the particles. And determining the fitness of the particles according to the optimization targets. And in response to determining that the number of iterative processing steps does not reach the preset iterative processing number, continuing to execute the iterative processing steps with the updated particles. And in response to determining that the number of iterative processing steps reaches a preset number of iterations, determining the updated particles obtained after the preset number of iterations as an optimal power scheduling scheme. And updating the optimal power scheduling scheme according to the fitness of the particles obtained after the preset iteration times to obtain an updated optimal power scheduling scheme. And then, superposing the set output curves corresponding to the updated optimal power dispatching scheme to obtain a superposed set output curve which is used as a set combined output curve. And finally, in response to determining that the shape similarity of the combined output curve of the unit and the curve of the load data of the predicted time period of the predicted day is higher than a preset similarity threshold, adjusting the power storage resources according to the updated optimal power scheduling scheme so as to realize collaborative scheduling of the virtual power plant. Therefore, the power loss generated by insufficient power supply or excessive power supply of the virtual power plant during power utilization can be reduced. Thus, stability of virtual power plant control is improved.
The above embodiments of the present disclosure have the following advantages: by the virtual power plant control method of some embodiments of the present disclosure, waste of power resources in the scheduling process is reduced. Specifically, the reason why the waste of the power resource is caused is that: because fluctuation law of load data change is difficult to predict, scheduling control resource allocation is unreasonable, when the scheduling control resource is overlarge, a power line with certain conveying capacity can not bear excessive electric energy, transmission loss is caused, and power resource waste is caused in the scheduling process of the power resource. Based on this, the virtual power plant control method of some embodiments of the present disclosure first obtains a historical daily load dataset and a predicted daily known period load data of the virtual power plant. Thus, a historical daily load data set of the virtual power plant and a predicted daily known period load data can be obtained for subsequent operation. And then, clustering the historical daily load data set and the predicted daily known period load data to obtain a clustering result. Therefore, the two data with higher similarity can be continuously combined until the clustering result is obtained after the combination. And then, according to the clustering result, a historical daily load data subset which has a similar relation with the load change rule after clustering between the load data in the known period of the predicted daily load and the historical daily load data set is screened. Thus, a subset of the historical daily load data that is the same class as the predicted daily known period load data can be selected from the clustering result so that the range similar to the predicted daily known period load data is narrowed. And then, according to the weather information and the predicted day type of the predicted day known period load data, carrying out data screening on the historical day load data subset to obtain a screened similar day load data set. Therefore, screening can be further carried out according to weather information and the predicted day type, a similar daily load data set after screening is obtained, and prediction errors in predicting load data of the virtual power plant are reduced. And then, generating a load data matrix according to the filtered similar daily load data set and the load data of the predicted daily known period. Thus, the filtered similar daily load data set and the predicted daily known period load data can be formed into a load data matrix. And then, inputting the load data matrix into a load data prediction model to obtain the load data of the prediction day prediction period. Therefore, the load data matrix can be predicted according to the load data prediction model, and the data which is not filled in the load data matrix can be filled, so that the load data of the prediction day prediction period can be obtained. And finally, carrying out cooperative scheduling control on the virtual power plant according to the forecast day forecast period load data. Therefore, the waste of the power resources in the scheduling process is reduced.
With further reference to FIG. 2, as an implementation of the method illustrated in the above figures, the present disclosure provides embodiments of a virtual power plant control method, which apparatus embodiments correspond to those illustrated in FIG. 1, and which apparatus is particularly applicable in a variety of electronic devices.
As shown in fig. 2, the virtual power plant control apparatus 200 of some embodiments includes: an acquisition unit 201, a clustering unit 202, a first screening unit 203, a second screening unit 204, a generation unit 205, an input unit 206, and a control unit 207. Wherein the obtaining unit 201 is configured to obtain a historical daily load data set and a predicted daily known period load data of the virtual power plant; a clustering unit 202 configured to cluster the historical daily load data set and the predicted daily known period load data to obtain a clustering result; a first filtering unit 203 configured to filter, according to the clustering result, a subset of historical daily load data having a similar relationship to the post-cluster load change rule between the load data of the known period of predicted daily load data from the set of historical daily load data; a second screening unit 204 configured to perform data screening on the historical daily load data subset according to weather information and the predicted daily type of the load data of the known period of time of the predicted day, so as to obtain a screened similar daily load data set; a generating unit 205 configured to generate a load data matrix based on the filtered similar daily load data set and the predicted daily known period load data; an input unit 206 configured to input the load data matrix to a load data prediction model to obtain predicted day prediction period load data; and a control unit 207 configured to perform cooperative scheduling control on the virtual power plant based on the predicted daily predicted period load data.
It will be appreciated that the elements described in the apparatus 200 correspond to the various steps in the method described with reference to fig. 1. Thus, the operations, features and resulting benefits described above for the method are equally applicable to the apparatus 200 and the units contained therein, and are not described in detail herein.
Referring now to FIG. 3, a schematic diagram of an electronic device (e.g., computing device) 300 suitable for use in implementing some embodiments of the present disclosure is shown. The electronic device shown in fig. 3 is merely an example and should not impose any limitations on the functionality and scope of use of embodiments of the present disclosure.
As shown in fig. 3, the electronic device 300 may include a processing means (e.g., a central processing unit, a graphics processor, etc.) 301 that may perform various suitable actions and processes in accordance with programs stored in a Read Only Memory (ROM) 302 or loaded from a storage 308 into a Random Access Memory (RAM) 304. In the RAM 303, various programs and data required for the operation of the electronic apparatus 300 are also stored. The processing device 301, the ROM 302, and the RAM 304 are connected to each other through a bus 304. An input/output (I/O) interface 305 is also connected to bus 304.
In general, the following devices may be connected to the I/O interface 305: input devices 306 including, for example, a touch screen, touchpad, keyboard, mouse, camera, microphone, accelerometer, gyroscope, etc.; an output device 307 including, for example, a Liquid Crystal Display (LCD), a speaker, a vibrator, and the like; storage 308 including, for example, magnetic tape, hard disk, etc.; and communication means 309. The communication means 309 may allow the electronic device 300 to communicate with other devices wirelessly or by wire to exchange data. While fig. 3 shows an electronic device 300 having various means, it is to be understood that not all of the illustrated means are required to be implemented or provided. More or fewer devices may be implemented or provided instead. Each block shown in fig. 3 may represent one device or a plurality of devices as needed.
In particular, according to some embodiments of the present disclosure, the processes described above with reference to flowcharts may be implemented as computer software programs. For example, some embodiments of the present disclosure include a computer program product comprising a computer program embodied on a computer readable medium, the computer program comprising program code for performing the method shown in the flow chart. In such embodiments, the computer program may be downloaded and installed from a network via communications device 309, or from storage device 308, or from ROM 302. The computer program, when executed by the processing means 301, performs the functions defined in the methods of some embodiments of the present disclosure.
It should be noted that, the computer readable medium described in some embodiments of the present disclosure may be a computer readable signal medium or a computer readable storage medium or any combination of the two. The computer readable storage medium can be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or a combination of any of the foregoing. More specific examples of the computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination. In some embodiments of the present disclosure, a computer readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. In some embodiments of the present disclosure, however, the computer-readable signal medium may comprise a data signal propagated in baseband or as part of a carrier wave, with the computer-readable program code embodied therein. Such a propagated data signal may take any of a variety of forms, including, but not limited to, electro-magnetic, optical, or any suitable combination thereof. A computer readable signal medium may also be any computer readable medium that is not a computer readable storage medium and that can communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device. Program code embodied on a computer readable medium may be transmitted using any appropriate medium, including but not limited to: electrical wires, fiber optic cables, RF (radio frequency), and the like, or any suitable combination.
In some implementations, the clients, servers may communicate using any currently known or future developed network protocol, such as HTTP (HyperText Transfer Protocol ), and may be interconnected with any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include a local area network ("LAN"), a wide area network ("WAN"), the internet (e.g., the internet), and peer-to-peer networks (e.g., ad hoc peer-to-peer networks), as well as any currently known or future developed networks.
The computer readable medium may be contained in the electronic device; or may exist alone without being incorporated into the electronic device. The computer readable medium carries one or more programs which, when executed by the electronic device, cause the electronic device to: acquiring a historical daily load data set and predicted daily known period load data of a virtual power plant; clustering the historical daily load data set and the predicted daily known period load data to obtain a clustering result; according to the clustering result, a historical daily load data subset which has a load change rule similarity relationship after clustering with the load data of the known period of the predicted daily load is screened from the historical daily load data set; according to the weather information and the predicted day type of the predicted day known period load data, carrying out data screening on the historical day load data subset to obtain a screened similar day load data set; generating a load data matrix according to the screened similar daily load data set and the load data of the predicted daily known period; inputting the load data matrix into a load data prediction model to obtain load data of a prediction day prediction period; and carrying out cooperative scheduling control on the virtual power plant according to the forecast day forecast period load data.
Computer program code for carrying out operations for some embodiments of the present disclosure may be written in one or more programming languages, including an object oriented programming language such as Java, python, C ++ and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code may execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any kind of network, including a Local Area Network (LAN) or a Wide Area Network (WAN), or may be connected to an external computer (for example, through the Internet using an Internet service provider).
The flowcharts and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present disclosure. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.
The units described in some embodiments of the present disclosure may be implemented by means of software, or may be implemented by means of hardware. The above described units may also be provided in a processor, for example, described as: a processor comprising: the device comprises an acquisition unit, a clustering unit, a first screening unit, a second screening unit, a generation unit, an input unit and a control unit. The names of these units do not constitute a limitation of the unit itself in some cases, and the acquisition unit may also be described as "a unit that acquires a historical daily load data set and predicted daily known period load data of a virtual power plant", for example.
The functions described above herein may be performed, at least in part, by one or more hardware logic components. For example, without limitation, exemplary types of hardware logic components that may be used include: a Field Programmable Gate Array (FPGA), an Application Specific Integrated Circuit (ASIC), an Application Specific Standard Product (ASSP), a system on a chip (SOC), a Complex Programmable Logic Device (CPLD), and the like.
The above description is only illustrative of some of the preferred embodiments of the present disclosure and of the principles of the technology employed above. It will be appreciated by those skilled in the art that the scope of the invention in question in the embodiments of the present disclosure is not limited to the specific combination of features described above, but encompasses other embodiments in which any combination of features described above or their equivalents is contemplated without departing from the inventive concepts described above. Such as the above-described features, are mutually substituted with (but not limited to) the features having similar functions disclosed in the embodiments of the present disclosure.
Claims (7)
1. A virtual power plant control method, comprising:
acquiring a historical daily load data set and predicted daily known period load data of a virtual power plant;
clustering the historical daily load data set and the predicted daily known period load data to obtain a clustering result;
according to the clustering result, a historical daily load data subset which has a similar relationship with the load change rule after clustering between the load data of the known period of the predicted day is screened out from the historical daily load data set;
according to weather information and predicted day types of the load data of the known period of the predicted day, carrying out data screening on the historical daily load data subset to obtain a screened similar daily load data set;
generating a load data matrix according to the screened similar daily load data set and the load data of the predicted daily known period;
inputting the load data matrix, the noise matrix and the mask matrix into a load data filling model in a load data prediction model to obtain a generating matrix;
filling the load data matrix according to the generation matrix to obtain an interpolation matrix, wherein load data corresponding to matrix elements in the interpolation matrix comprises real data and false data;
Generating load data of a prediction day prediction period according to the interpolation matrix;
defining a load data noise matrix for each load data sample in the load data sample set, wherein the load data matrix and the mask matrix are load data matrix sample sets, and a load data matrix sample set is obtained;
defining a prompt matrix aiming at a mask matrix of each load data sample in the load data sample set to obtain a prompt matrix set;
for each load data matrix sample set of the set of load data matrix sample sets, performing the following first processing step:
inputting the load data noise matrix in the load data matrix sample group, the load data matrix and the mask matrix into a load data filling model in the load data prediction model to obtain a load data interpolation matrix;
inputting the load data interpolation matrix and the prompt matrix in the prompt matrix set into a load data probability judging model in the load data prediction model to obtain a predicted mask matrix, wherein each element in the predicted mask matrix represents the probability of whether load data is real data or not, and the prompt matrix is a prompt matrix of a load data sample which is the same as the load data mask matrix in the prompt matrix set;
In response to determining that the load data interpolation matrix is not equal to the load data matrix, adjusting network parameters of the load data padding model;
adjusting network parameters of the load data probability discrimination model in response to determining that the predicted mask matrix is not equal to the mask matrix;
determining the load data filling model as a final load data filling model in response to determining that the difference between the data in the load data interpolation matrix and the data corresponding to the load data matrix is less than a preset first threshold;
determining a load data probability discrimination model as a final load data probability discrimination model in response to determining that the difference between the predicted mask matrix and the mask matrix is less than a preset second threshold;
determining the final load data filling model and the final load data probability judging model as an optimized load data prediction model;
and carrying out cooperative scheduling control on the virtual power plant according to the load data of the forecast day forecast period.
2. The method of claim 1, wherein the clustering the historical daily load dataset and the predicted daily known period load data to obtain a clustered result comprises:
Determining the load data of the known period of the predicted day and the historical daily load data set as standardized post-load data sets;
the following clustering steps are performed on the normalized post-load dataset:
determining the similarity between every two standardized load data in the standardized load data set to obtain a similarity set;
determining the first standardized post-load data and the second standardized post-load data corresponding to the maximum similarity in the similarity set as standardized post-load data types;
deleting the first normalized post-load data and the second normalized post-load data from the normalized post-load data set to obtain a deleted post-load data set;
determining each deleted load data in the deleted load data set as a deleted load data class, and obtaining a deleted load data class set;
combining the standardized load data class and the deleted load data class set into a clustered load data class set;
in response to determining that the number of clustered load data classes in the clustered load data class set is greater than or equal to a preset clustering number, determining the clustered load data class set as a standardized load data set, and executing the clustering step again;
And determining the obtained clustered load data class set as a clustering result in response to determining that the number of clustered load data classes in the clustered load data class set is smaller than the preset clustering number.
3. The method according to claim 1, wherein the data filtering the subset of historical daily load data according to weather information and predicted daily type of the predicted daily known period load data to obtain a filtered similar daily load data set comprises:
according to the weather information and the predicted day type, determining the data similarity between each historical day load data in the historical day load data subset and the load data of the known period of the predicted day to obtain a data similarity set;
and for each data similarity in the data similarity set, determining historical daily load data corresponding to the data similarity larger than the preset data similarity as the screened similar daily load data in response to determining that the data similarity is larger than the preset data similarity, and obtaining a screened similar daily load data set.
4. The method of claim 1, wherein the co-scheduling the virtual power plant based on the predicted daily predicted period load data comprises:
In response to determining that a load change curve corresponding to the load data of the forecast day forecast period is an ascending curve, increasing the power generation amount of the virtual power plant unit so as to perform cooperative scheduling control on the virtual power plant;
and in response to determining that the load change curve corresponding to the load data of the forecast day forecast period is a descending curve, reducing the power generation amount of the virtual power plant unit so as to perform cooperative scheduling control on the virtual power plant.
5. A virtual power plant control apparatus, comprising:
an acquisition unit configured to acquire a historical daily load data set and predicted daily known period load data of the virtual power plant;
the clustering unit is configured to cluster the historical daily load data set and the predicted daily known period load data to obtain a clustering result;
a first screening unit configured to screen, according to the clustering result, a historical daily load data subset having a similar relationship with a post-cluster load change rule between the load data of the known period of the predicted day from the historical daily load data set;
the second screening unit is configured to perform data screening on the historical daily load data subset according to weather information and the predicted daily type of the load data in the known period of the predicted day to obtain a screened similar daily load data set;
A generation unit configured to generate a load data matrix from the filtered similar daily load data set and the predicted daily known period load data;
the input unit is configured to input the load data matrix, the noise matrix and the mask matrix into a load data filling model in a load data prediction model to obtain a generation matrix; filling the load data matrix according to the generation matrix to obtain an interpolation matrix, wherein load data corresponding to matrix elements in the interpolation matrix comprises real data and false data; generating load data of a prediction day prediction period according to the interpolation matrix; defining a load data noise matrix for each load data sample in the load data sample set, wherein the load data matrix and the mask matrix are load data matrix sample sets, and a load data matrix sample set is obtained; defining a prompt matrix aiming at a mask matrix of each load data sample in the load data sample set to obtain a prompt matrix set; for each load data matrix sample set of the set of load data matrix sample sets, performing the following first processing step: inputting the load data noise matrix in the load data matrix sample group, the load data matrix and the mask matrix into a load data filling model in the load data prediction model to obtain a load data interpolation matrix; inputting the load data interpolation matrix and the prompt matrix in the prompt matrix set into a load data probability judging model in the load data prediction model to obtain a predicted mask matrix, wherein each element in the predicted mask matrix represents the probability of whether load data is real data or not, and the prompt matrix is a prompt matrix of a load data sample which is the same as the load data mask matrix in the prompt matrix set; in response to determining that the load data interpolation matrix is not equal to the load data matrix, adjusting network parameters of the load data padding model; adjusting network parameters of the load data probability discrimination model in response to determining that the predicted mask matrix is not equal to the mask matrix; determining the load data filling model as a final load data filling model in response to determining that the difference between the data in the load data interpolation matrix and the data corresponding to the load data matrix is less than a preset first threshold; determining a load data probability discrimination model as a final load data probability discrimination model in response to determining that the difference between the predicted mask matrix and the mask matrix is less than a preset second threshold; determining the final load data filling model and the final load data probability judging model as an optimized load data prediction model;
And the control unit is configured to perform cooperative scheduling control on the virtual power plant according to the forecast day forecast period load data.
6. An electronic device, comprising:
one or more processors;
a storage device having one or more programs stored thereon;
when executed by the one or more processors, causes the one or more processors to implement the method of any of claims 1 to 4.
7. A computer readable medium having stored thereon a computer program, wherein the program when executed by a processor implements the method of any of claims 1 to 4.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202311147299.9A CN116882597B (en) | 2023-09-07 | 2023-09-07 | Virtual power plant control method, device, electronic equipment and readable medium |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202311147299.9A CN116882597B (en) | 2023-09-07 | 2023-09-07 | Virtual power plant control method, device, electronic equipment and readable medium |
Publications (2)
Publication Number | Publication Date |
---|---|
CN116882597A CN116882597A (en) | 2023-10-13 |
CN116882597B true CN116882597B (en) | 2024-01-12 |
Family
ID=88272111
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202311147299.9A Active CN116882597B (en) | 2023-09-07 | 2023-09-07 | Virtual power plant control method, device, electronic equipment and readable medium |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN116882597B (en) |
Families Citing this family (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117332236B (en) * | 2023-10-30 | 2024-05-07 | 国网浙江省电力有限公司乐清市供电公司 | Data tracking detection method, device and storage medium for virtual power plant |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113505943A (en) * | 2021-07-30 | 2021-10-15 | 广东电网有限责任公司 | Method, system, equipment and medium for predicting short-term load of power grid |
CN113852135A (en) * | 2021-09-22 | 2021-12-28 | 广东电网有限责任公司 | Virtual power plant energy scheduling method, device, storage medium and platform |
CN115085196A (en) * | 2022-08-19 | 2022-09-20 | 国网信息通信产业集团有限公司 | Power load predicted value determination method, device, equipment and computer readable medium |
Family Cites Families (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20170140278A1 (en) * | 2015-11-18 | 2017-05-18 | Ca, Inc. | Using machine learning to predict big data environment performance |
-
2023
- 2023-09-07 CN CN202311147299.9A patent/CN116882597B/en active Active
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113505943A (en) * | 2021-07-30 | 2021-10-15 | 广东电网有限责任公司 | Method, system, equipment and medium for predicting short-term load of power grid |
CN113852135A (en) * | 2021-09-22 | 2021-12-28 | 广东电网有限责任公司 | Virtual power plant energy scheduling method, device, storage medium and platform |
CN115085196A (en) * | 2022-08-19 | 2022-09-20 | 国网信息通信产业集团有限公司 | Power load predicted value determination method, device, equipment and computer readable medium |
Non-Patent Citations (1)
Title |
---|
考虑气象因素的短期负荷预测模型研究;朱祥和;;数学的实践与认识(03);全文 * |
Also Published As
Publication number | Publication date |
---|---|
CN116882597A (en) | 2023-10-13 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN116882597B (en) | Virtual power plant control method, device, electronic equipment and readable medium | |
CN110232471A (en) | A kind of precipitation sensing net node layout optimization method and device | |
CN109886699A (en) | Activity recognition method and device, electronic equipment, storage medium | |
CN112308281A (en) | Temperature information prediction method and device | |
CN113807589B (en) | Rolling optimization method and device for energy storage of operators based on model prediction | |
CN116703131B (en) | Power resource allocation method, device, electronic equipment and computer readable medium | |
CN117117833A (en) | Photovoltaic output power prediction method and device, electronic equipment and storage medium | |
CN115759413A (en) | Meteorological prediction method and device, storage medium and electronic equipment | |
CN117236805B (en) | Power equipment control method, device, electronic equipment and computer readable medium | |
CN115907136B (en) | Electric automobile dispatching method, device, equipment and computer readable medium | |
CN115360717B (en) | Power distribution network scheduling method, system, equipment and medium for comprehensive energy | |
CN112884189A (en) | Order quantity prediction model training method, device and equipment | |
JP2023083276A (en) | Method and device for training information adjustment model of charging station, and product | |
CN113723712B (en) | Wind power prediction method, system, equipment and medium | |
CN115730812A (en) | Index information generation method and device, electronic equipment and readable medium | |
CN115187144A (en) | Virtual power plant power flow method, device, equipment, medium and program product | |
Aliberti et al. | Non-linear autoregressive neural networks to forecast short-term solar radiation for photovoltaic energy predictions | |
CN117235535B (en) | Abnormal supply end power-off method and device, electronic equipment and medium | |
CN116911577B (en) | Comprehensive energy scheduling method, device, electronic equipment and computer readable medium | |
CN116862319B (en) | Power index information generation method, device, electronic equipment and medium | |
CN115577980B (en) | Power equipment regulation and control method and device, electronic equipment and medium | |
CN117713048A (en) | Power supply method, device and equipment for power utilization terminal based on rural power network power utilization integration | |
CN116451168B (en) | Abnormal power information generation method, device, electronic equipment and readable medium | |
CN117196199A (en) | Article scheduling method, apparatus, electronic device and computer readable medium | |
CN117913779A (en) | Method, apparatus, electronic device and readable medium for predicting electric load information |
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 | ||
GR01 | Patent grant | ||
GR01 | Patent grant |