US-12619805-B2 - Method for planning a layout of a renewable energy site

US12619805B2US 12619805 B2US12619805 B2US 12619805B2US-12619805-B2

Abstract

Correlated sets of historical meteorological data and terrain data are obtained for at least one geographical area. A data model is derived based on the basis of the correlated sets, by training the data model. The trained data model is adapted to identify coherence between meteorological data and terrain data relating to the same geographical area. Meteorological data and terrain data related to the renewable energy site are fed to the trained data model, the terrain data having a higher resolution than the meteorological data. Using the trained model, meteorological data related to the renewable energy site with increased resolution is estimated by downscaling the meteorological data. The estimated meteorological data with increased resolution for the renewable energy site is then used for planning a layout of the renewable energy site.

Inventors

Martin QVIST
Ana Maria Martinez FERNANDEZ
Hans Harhoff ANDERSEN
Hjalte Vinther KIEFER

Assignees

VESTAS WIND SYSTEMS A/S

Dates

Publication Date: 20260505
Application Date: 20200914
Priority Date: 20191003

Claims (20)

1 . A method for planning a layout of a renewable energy site, the method comprising: obtaining, for at least one geographical area, correlated sets of historical meteorological data and terrain data, relating to the respective geographical area(s); deriving a data model using a deep learning algorithm, and on the basis of the correlated sets of historical meteorological data and terrain data, by training the data model, the data model being adapted to identify coherence between meteorological data and terrain data relating to the same geographical area; feeding site-specific meteorological data and site-specific terrain data related to the renewable energy site to the data model, the site-specific terrain data having a higher resolution than the site-specific meteorological data; obtaining estimated meteorological data related to the renewable energy site with increased resolution by downscaling the site-specific meteorological data, using the trained data model, and based on the site-specific meteorological data and the site-specific terrain data fed to the data model; planning the layout of the renewable energy site on the basis of the estimated meteorological data related to the renewable energy site with increased resolution; and constructing the renewable energy site in accordance with the layout of the renewable energy site.
2 . The method of claim 1 , wherein the obtaining terrain data comprises obtaining elevation data.
3 . The method of claim 1 , wherein the obtaining terrain data comprises obtaining roughness data.
4 . The method of claim 1 , wherein the historical meteorological data comprises high resolution historical meteorological data.
5 . The method of claim 1 , wherein the terrain data comprises high resolution terrain data.
6 . The method of claim 1 , wherein the historical meteorological data is in the form of one or more time series.
7 . The method of claim 1 , further comprising estimating an energy production of the renewable energy site based on the estimated meteorological data with increased resolution for the renewable energy site and based on the layout of the renewable energy site.
8 . The method of claim 1 , further comprising: operating the renewable energy site for a predefined time period; obtaining additional meteorological data during the renewable energy site during the predefined time period; feeding the additional meteorological data obtained during the predefined time period to the data model; obtaining additional estimated meteorological data related to the renewable energy site and to the predefined time period with increased resolution by downscaling the additional meteorological data using the data model, and based on the site-specific terrain data previously fed to the data model; and estimating an expected energy production of the renewable energy site for the predefined time period, based on the additional estimated meteorological data with increased resolution for the renewable energy site for the predefined time period.
9 . The method of claim 1 , wherein the renewable energy site comprises at least one wind turbine generator.
10 . The method of claim 1 , wherein the deep learning algorithm comprises one or more help functions for deriving the data model, the one or more help functions being adapted to provide information regarding a vector field defined by the historical meteorological data.
11 . The method of claim 1 , wherein the meteorological data comprises wind data and/or solar influx data.
12 . The method of claim 1 , wherein the deep learning algorithm comprises convolutional network models, recurrent neural networks, generative adversarial network models and/or feed forward models.
13 . The method of claim 1 , wherein the correlated sets of historical meteorological data and terrain data are obtained from at least two different geographical areas.
14 . A method for identifying a renewable energy site, the method comprising: obtaining, for at least one geographical area, correlated sets of historical meteorological data and terrain data, relating to the respective geographical area(s); deriving a data model using a deep learning algorithm, and on the basis of the correlated sets of historical meteorological data and terrain data, by training the data model, the data model being adapted to identify coherence between meteorological data and terrain data relating to the same geographical area; feeding site-specific meteorological data and site-specific terrain data related to at least one geographical area to the data model, each geographical area comprising at least one candidate renewable energy site, the site-specific terrain data having a higher resolution than the site-specific meteorological data; obtaining estimated meteorological data related to the geographical area(s) with increased resolution by downscaling the site-specific meteorological data, using the data model, and based on the site-specific meteorological data and the site-specific terrain data fed to the data model; identifying at least one suitable renewable energy site within the at least one geographical area on the basis of the estimated meteorological data related to the geographical area(s) with increased resolution; and estimating an energy production of the renewable energy site based on the estimated meteorological data with increased resolution for the renewable energy site and based on the layout of the renewable energy site.
15 . The method of claim 14 , wherein the obtaining terrain data comprises obtaining at least one of elevation data or roughness data.
16 . The method of claim 14 , wherein the terrain data comprises high resolution terrain data.
17 . The method of claim 14 , wherein the historical meteorological data comprises high resolution historical meteorological data.
18 . The method of claim 14 , wherein the historical meteorological data is in the form of one or more time series.
19 . The method of claim 14 , further comprising: constructing the renewable energy site in accordance with the layout of the renewable energy site.
20 . The method of claim 19 , further comprising: operating the renewable energy site for a predefined time period; obtaining additional meteorological data during the renewable energy site during the predefined time period; feeding the additional meteorological data obtained during the predefined time period to the data model; obtaining additional estimated meteorological data related to the renewable energy site and to the predefined time period with increased resolution by downscaling the additional meteorological data using the data model, and based on the site-specific terrain data previously fed to the data model; and estimating an expected energy production of the renewable energy site for the predefined time period, based on the additional estimated meteorological data with increased resolution for the renewable energy site for the predefined time period.

Description

FIELD OF THE INVENTION The present invention relates to a method for planning a layout of a renewable energy site, such as a site of a wind energy plant, using a deep learning algorithm. BACKGROUND OF THE INVENTION Knowing the operational conditions, such as wind-related conditions, solar conditions, etc., for an area of interest during planning of a renewable energy site, such as a wind energy plant or a solar energy plant, is of maximum relevance. In some cases, such operational conditions have previously been obtained by means of simulations, e.g. performed using large data sets of relevant historical information regarding the operational conditions. The simulations could, e.g., include downscaling of the data. However, such simulations are quite computationally expensive and may take several hours to be completed. The more data to be processed, the more time consuming simulations may be expected to be, and this may further increase the needs for computing power. Factors such as the complexity of the terrain of the area of interest and/or the target resolution of the downscaled data may further increase the simulation time. One approach for downscaling large data sets is referred to as dynamical downscaling. Dynamical downscaling requires running high-resolution climate models on a regional sub-domain, using observational data or lower-resolution climate model output as a boundary condition. These models use physical principles to reproduce local climates, but are computationally expensive. Another approach is downscaling using statistical methods. Statistical downscaling is a two-step process in which statistical relationships between local climate variables, such as surface aft temperature, precipitation, etc., and large-scale predictors, such as pressure fields, are initially developed, and the relationships are subsequently applied to outputs of global climate model experiments in order to simulate local climate characteristics in the future. This approach is computationally less expensive than the dynamical downscaling approach, but is not reliable, since it can not guarantee a reasonable result for every simulation. DESCRIPTION OF THE INVENTION It is an object of embodiments of the invention to provide a method for planning a layout of a renewable energy site in a fast and reliable manner. It is further an object of embodiments of the invention to provide a method for planning a layout of a renewable energy site in a computationally inexpensive manner. According to a first aspect the invention provides a method for planning a layout of a renewable energy site, the method comprising the steps of: obtaining, for at least one geographical area, correlated sets of historical meteorological data and terrain data, relating to the respective geographical area(s),deriving a data model using a deep learning algorithm, and on the basis of the correlated sets of historical meteorological data and terrain data, by training the data model, the trained data model being adapted to identify coherence between meteorological data and terrain data relating to the same geographical area,feeding meteorological data and terrain data related to the renewable energy site to the trained data model, the terrain data having a higher resolution than the meteorological data,obtaining estimated meteorological data related to the renewable energy site with increased resolution by downscaling the meteorological data, using the trained data model, and based on the data fed to the trained data model, andplanning the layout of the renewable energy site on the basis of the estimated meteorological data with increased resolution for the renewable energy site. Thus, according to the first aspect, the invention provides a method for planning a layout of a renewable energy site. In the present context the term ‘renewable energy site’ should be interpreted to mean a collection of two or more renewable energy generating units, such as wind turbines, photovoltaic cells, etc., arranged within a limited geographical area, and which may share various forms of infrastructure, such as access roads, communication network, substations, power electronics, grid connections, etc. In the present context, the term ‘renewable energy generating’ unit should be interpreted to mean a unit which is capable of generating electrical power, and which supplies all or part of the produced power to a power grid. In the method according to the first aspect of the invention, correlated sets of historical meteorological data and terrain data, relating to at least one geographical area are initially obtained, Thus, for each geographical area, information regarding historical meteorological conditions occurring in that geographical area, as well as information regarding terrain in that geographical area is obtained. For a given geographical area, the historical meteorological data thereby forms one part of the correlated set of data and the terrain data forms the othe