KR-102961225-B1 - System for prediction of wind power generation, Method, and Applications using the same

KR102961225B1KR 102961225 B1KR102961225 B1KR 102961225B1KR-102961225-B1

Abstract

A wind power generation prediction system capable of predicting wind power generation in a wind power generation complex that generates wind power using multiple wind turbines comprises: a wind power generation complex relationship diagram derivation unit that derives a wind power generation complex relationship diagram, which is a diagram of relationships between wind turbines within a wind power generation complex, wherein wind turbines are nodes and the spatial relationships between wind turbines are edges; a prediction information output unit that obtains the wind power generation complex relationship diagram including time series information, learns multiple wind power generation complex relationship diagrams, and outputs prediction information corresponding to a prediction condition when a specific prediction condition is input; and a learning correction unit that compares the prediction information with actual measurement information, and if the comparison result shows that the actual measurement information differs from the prediction information, modifies the prediction information to the actual measurement information to perform a correction of the learning.

Inventors

김웅경
송석일

Assignees

국립한국교통대학교산학협력단

Dates

Publication Date: 20260507
Application Date: 20231213

Claims (13)

In a wind power generation prediction system capable of predicting wind power generation in a wind power complex that generates wind power using multiple wind turbines, The above wind power generation prediction system comprises: a wind power generation complex relationship diagram derivation unit that derives a wind power generation complex relationship diagram, which is a relationship diagram between wind turbines within the wind power generation complex, wherein the wind turbines are nodes and the spatial relationships between the wind turbines are edges; The above wind power plant relationship diagram is obtained including time series information, and a prediction information output unit learns a plurality of the above wind power plant relationship diagrams and outputs prediction information corresponding to the prediction condition when a specific prediction condition is input; and A wind power generation prediction system comprising: a learning correction unit that compares the above-mentioned prediction information with actual measurement information, and if the comparison result shows that the above-mentioned actual measurement information is different from the above-mentioned prediction information, modifies the above-mentioned prediction information to the above-mentioned actual measurement information to perform correction of the learning.
In Article 1, A wind power generation prediction system wherein the above node refers to turbine information obtained from the wind turbine, and the turbine information includes at least one of active power, reactive power, wind speed, wind direction, temperature, turbine nacelle direction, turbine nacelle internal temperature, blade pitch angle, and the absolute position of the turbine or the relative position within a preset analysis area measured at the wind turbine.
In Article 1, A wind power generation prediction system in which the above edge is formed to connect adjacent nodes, and the edge includes distance and angle information between the connected adjacent nodes.
In Article 1, The above prediction information output unit is, A wind power generation prediction system that learns the above wind power complex relationship diagram to obtain feature vectors, groups the obtained feature vectors into a preset time range, and converts them into range feature vectors.
In Paragraph 4, The above prediction information output unit is, A wind power generation prediction system that processes the prediction conditions using the above-mentioned range feature vector and outputs the effective power corresponding to the prediction conditions as the prediction information through the processing result.
In Article 1, It further includes an optimal installation location derivation unit that derives an optimal installation location using a prediction information derivation algorithm, which is an algorithm used in the wind power complex relationship diagram and the prediction information output unit, to actually or virtually install additional wind turbines; The above optimal installation location derivation unit is, Additional wind turbine information acquisition module for acquiring additional wind turbine information, which is information of the additional wind turbine mentioned above; An installation location candidate derivation module that obtains the above prediction information derivation algorithm and derives the additional wind turbine installation location candidate by applying the additional wind turbine information to the above prediction information derivation algorithm; and A wind power generation prediction system comprising: an installation location determination module that sets the additional wind turbine as a virtual node in the above installation location candidates, derives virtual edge information connecting the virtual node and other nodes to obtain a virtual wind power complex relationship diagram, and determines the optimal additional wind turbine installation location using the virtual wind power complex relationship diagram.
In a wind power generation prediction method capable of predicting wind power generation in a wind power complex that generates wind power using multiple wind turbines, The above wind power generation prediction method comprises a step of deriving a wind power complex relationship diagram, which is a relationship diagram between wind turbines within the wind power complex, wherein the wind turbines are nodes and the spatial relationships between the wind turbines are edges; The above wind power plant relationship diagram is obtained including time series information, and a prediction information output step of learning a plurality of the above wind power plant relationship diagrams and, when a specific prediction condition is input, outputting prediction information corresponding to the prediction condition; and A wind power generation prediction method comprising: a learning correction step of comparing the above-mentioned prediction information with actual measurement information, and if the comparison result shows that the above-mentioned actual measurement information is different from the above-mentioned prediction information, modifying the above-mentioned prediction information with the above-mentioned actual measurement information to perform correction of the learning.
In Article 7, The above node is, A wind power generation prediction method comprising: turbine information obtained from the wind turbine, wherein the turbine information includes at least one of active power, reactive power, wind speed, wind direction, temperature, turbine nacelle direction, turbine nacelle internal temperature, blade pitch angle, and the absolute position of the turbine or the relative position within a preset analysis area measured at the wind turbine.
In Article 7, The above edge is, A wind power generation prediction method formed to connect adjacent nodes, wherein the edge includes distance and angle information between the connected adjacent nodes.
In Article 7, The above prediction information output step is, A wind power generation prediction method that learns the above wind power complex relationship diagram to obtain feature vectors, groups the obtained feature vectors into a preset time range, and converts them into range feature vectors.
In Article 10, The above prediction information output step is, A wind power generation prediction method that processes the prediction conditions using the above-mentioned range feature vector and outputs the effective power corresponding to the prediction conditions as the prediction information through the processing result.
In Article 7, It further includes an optimal installation location derivation step for deriving an optimal installation location using a prediction information derivation algorithm, which is an algorithm used in the wind power complex relationship diagram and the prediction information output step, to actually or virtually install additional wind turbines; The above step of deriving the optimal installation location is, A step for acquiring additional wind turbine information, which is information of the additional wind turbine mentioned above; An installation location candidate derivation step of obtaining the above-mentioned prediction information derivation algorithm and deriving the above-mentioned additional wind turbine installation location candidate by applying the above-mentioned additional wind turbine information to the above-mentioned prediction information derivation algorithm; and A wind power generation prediction method comprising: an installation location determination step of setting the additional wind turbine as a virtual node in the above installation location candidates, deriving virtual edge information connecting the virtual node and other nodes to obtain a virtual wind power complex relationship diagram, and determining the optimal additional wind turbine installation location using the virtual wind power complex relationship diagram.
In an application using a wind power generation prediction method capable of predicting wind power generation in a wind power complex that generates wind power using multiple wind turbines, A step for deriving a relationship diagram of a wind power complex, wherein the wind turbines are nodes and the spatial relationships between the wind turbines are edges; The above wind power plant relationship diagram is obtained including time series information, and a prediction information output step of learning a plurality of the above wind power plant relationship diagrams and, when a specific prediction condition is input, outputting prediction information corresponding to the prediction condition; and An application stored in a digital terminal to perform a learning correction step of comparing the above-mentioned prediction information and actual measurement information, and if the comparison result shows that the above-mentioned actual measurement information is different from the above-mentioned prediction information, modifying the above-mentioned prediction information to the above-mentioned actual measurement information to perform the learning correction.

Description

System for prediction of wind power generation, Method, and Applications using the same The present invention relates to a wind power generation prediction system, a method, and an application using the same. In particular, it relates to a wind power generation prediction system, a method, and an application using the same that can predict the output of a wind turbine by acquiring and analyzing relational information of wind turbines in a wind power generation complex. A major issue regarding wind power generation facilities is that wind power generation is dependent on wind direction and speed information. Since the output of wind power varies depending on wind intensity and frequency, energy cannot be produced on windless days. Furthermore, if the wind is excessively strong and poses a risk of damage, the system must be shut down to protect the facility, which similarly results in a lack of energy production. The Jeju region is an example of such a wind power complex. There are numerous wind power complexes in the Jeju region, and output and capacity limits for wind turbines are increasing to ensure the stability of the power grid. Looking at the case of 2020, power output was restricted 77 times in the Jeju region. Predicting wind power generation is a critical factor in energy management and power grid stability. If accurate wind power forecasting is possible, it offers the advantages of more efficient energy supply management, preparation of reserve generation capacity, and maintenance of power system stability. However, wind power generation is not predicted simply by using wind speed, wind direction, and frequency alone; as mentioned above, in the case of large-scale wind power complexes, the output can vary depending on the installation location of the turbines. This is because, as it is a large-scale wind power complex, there may be differences in wind direction and speed depending on the installation location of the wind turbines. Furthermore, since wind turbines are installed in multiple rows rather than in a single line, the wind speed of the preceding or adjacent wind turbines decreases and the turbulence intensity increases due to the wake effect of surrounding turbines, resulting in a decrease in power output. Consequently, there is a problem in that accurate prediction is difficult using simple wind power prediction methods. FIG. 1 is a block diagram of a wind power generation prediction system according to an embodiment of the present invention. FIG. 2 is a block diagram of an optimal installation location derivation unit according to an embodiment of the present invention. FIG. 3 is a flowchart of a wind power generation prediction method according to an embodiment of the present invention. Figure 4 is a flowchart of step S17 of Figure 3. FIG. 5 is a diagram of a wind power plant according to one embodiment of the present invention. FIG. 6 is an example diagram of performing learning and outputting prediction information in one embodiment of the present invention. Hereinafter, some embodiments of the present disclosure will be described in detail with reference to the exemplary drawings. In assigning reference numerals to the components of each drawing, the same components may have the same reference numeral as much as possible, even if they are shown in different drawings. Furthermore, in describing the embodiments, if it is determined that a detailed description of related known components or functions may obscure the essence of the technical concept, such detailed description may be omitted. Where terms such as "comprising," "having," or "consisting of" are used in this specification, other parts may be added unless "only" is used. Where a component is expressed in the singular, it may include a plural unless otherwise specified. Additionally, terms such as first, second, A, B, (a), (b), etc., may be used to describe the components of the present disclosure. These terms are used merely to distinguish the components from other components, and the nature, order, sequence, or number of the components are not limited by such terms. In describing the positional relationship of components, where it is stated that two or more components are "connected," "combined," or "joined," it should be understood that while the two or more components may be directly "connected," "combined," or "joined," they may also be "connected," "combined," or "joined" with other components "intervened." Here, the other components may be included in one or more of the two or more components that are "connected," "combined," or "joined" with one another. In describing the temporal flow relationship regarding components, methods of operation, or methods of production, for example, when the temporal or sequential relationship is described using "after," "following," "next," or "before," it may include cases where the relationship is not continuous unless "immediately" or "directly" is used. Meanwhile, where numerical val