Package {windfarmGA}

windfarmGA: Genetic Algorithm for Wind Farm Layout Optimization

Description

The genetic algorithm is designed to optimize wind farms of any shape. Each layout is encoded as n unique grid-cell identifiers. It requires a predefined amount of turbines, a unified rotor radius and an average wind speed value for each incoming wind direction. A terrain effect model can be included that downloads an 'SRTM' elevation model and loads a Corine Land Cover raster to approximate surface roughness.

Details

A package to optimize small wind farms with irregular shapes using a genetic algorithm. Each individual is n unique grid-cell IDs (set-crossover, swap-mutation). It requires a fixed amount of turbines, a fixed rotor radius and an average wind speed value for each incoming wind direction. A terrain effect model can be included which downloads a digital elevation model and a Corine Land Cover raster to approximate surface roughness. Further information can be found at the description of the function genetic_algorithm.

Author(s)

Maintainer: Sebastian Gatscha sebastian_gatscha@gmx.at

See Also

Useful links:

• Documentation Github.io

• Documentation

• Master Thesis

• Shiny App

• Report Issues

Mark a genetic_algorithm result

Description

Attach the windfarmGA class so print() and plot() dispatch. New runs from genetic_algorithm() already have this class. print() shows run inputs, the wind table, and each generation that set a new fitness maximum.

Usage

as_windfarmGA(x)

## S3 method for class 'windfarmGA'
print(x, ...)

## S3 method for class 'windfarmGA'
plot(x, y, ...)

Arguments

Value

x with class windfarmGA.

Calculates Air Density, Air Pressure and Temperature according to the Barometric Height Formula

Description

Calculates air density, temperature and air pressure respective to certain heights according to the International standard atmosphere and the barometric height formula.

Usage

barometric_height(data, height, po = 101325, ro = 1.225)

Arguments

Value

Returns a data.frame with height values and corresponding air pressures, air densities and temperatures in Kelvin and Celsius.

See Also

Other Wind Energy Calculation Functions: calculate_energy(), circle_intersection(), get_dist_angles(), turbine_influences()

Examples

data <- matrix(seq(0, 5000, 500))
barometric_height(data)
plot.ts(barometric_height(data))

A POLYGON with an area of ~70 km2

Description

A POLYGON with an area of ~70 km2

Usage

big_shape

Format

An object of class sf (inherits from data.frame) with 1 rows and 1 columns.

Calculate Energy Outputs of Individuals

Description

Calculate the energy output and efficiency rates of an individual in the current population under all given wind directions and speeds. If the terrain effect model is activated, the main calculations to model those effects will be done in this function.

Usage

calculate_energy(
  layout,
  reference_height,
  rotor_height,
  surface_roughness,
  wake_angle,
  wake_distance,
  area,
  rotor,
  wind,
  elevation = NULL,
  terrain = FALSE,
  ccl_raster = NULL,
  weibull = FALSE,
  park_center = NULL,
  plot = FALSE
)

Arguments

Value

Returns a list of an individual of the current generation with resulting wake effects, energy outputs, efficiency rates for every wind direction. The length of the list corresponds to the number of different wind directions.

See Also

Other Wind Energy Calculation Functions: barometric_height(), circle_intersection(), get_dist_angles(), turbine_influences()

Examples

## Create a random Polygon
library(sf)
area <- sf::st_as_sf(sf::st_sfc(
  sf::st_polygon(list(cbind(
    c(4498482, 4498482, 4499991, 4499991, 4498482),
    c(2668272, 2669343, 2669343, 2668272, 2668272)
  ))),
  crs = 3035
))

## Create a uniform and unidirectional wind data.frame and plot the
## resulting wind rose
data.in <- data.frame(ws = 12, wd = 0)
windrosePlot <- plot_windrose(
  data = data.in, spd = data.in$ws,
  dir = data.in$wd, dirres = 10, spdmax = 20
)

## Assign the rotor radius and a factor of the radius for grid spacing.
rotor <- 50
fcr <- 3
resGrid <- grid_area(
  area = area, size = rotor * fcr, prop = 1,
  plot_grid = TRUE
)
## Assign the indexed data frame to new variable. Element 2 of the list
## is the grid, saved as Simple Feature Polygons.
resGrid1 <- resGrid[[1]]

## Create an initial population with the indexed Grid, 15 turbines and
## 100 individuals.
initpop <- init_population(grid = resGrid1, n = 15, n_start = 100)

## Calculate the expected energy output of the first individual of the
## population.
par(mfrow = c(1, 2))
plot(area)
points(initpop[[1]][, "X"], initpop[[1]][, "Y"], pch = 20, cex = 2)
plot(resGrid[[2]], add = TRUE)
resCalcEn <- calculate_energy(
  layout = initpop[[1]], reference_height = 50,
  rotor_height = 50, surface_roughness = 0.14, wake_angle = 20,
  wake_distance = 100000, wind = data.in,
  rotor = 50, area = area, terrain = FALSE,
  weibull = FALSE
)
resCalcEn <- as.data.frame(resCalcEn)
plot(area, main = resCalcEn[, "Energy_Output_Red"][[1]])
points(x = resCalcEn[, "Bx"], y = resCalcEn[, "By"], pch = 20)


## Create a variable and multidirectional wind data.frame and plot the
## resulting wind rose
data.in10 <- data.frame(ws = runif(10, 1, 25), wd = runif(10, 0, 360))
windrosePlot <- plot_windrose(
  data = data.in10, spd = data.in10$ws,
  dir = data.in10$wd, dirres = 10, spdmax = 20
)

## Calculate the energy outputs for the first individual with more than one
## wind direction.
resCalcEn <- calculate_energy(
  layout = initpop[[1]], reference_height = 50,
  rotor_height = 50, surface_roughness = 0.14, wake_angle = 20,
  wake_distance = 100000, wind = data.in10,
  rotor = 50, area = area, terrain = FALSE,
  weibull = FALSE
)

Get area of intersecting circles

Description

Calculate the intersection area of two circles with different radii and different heights

Usage

circle_intersection(r1, r2, h1, h2, dx)

Arguments

Value

A numeric vector; one intersection area per pair. Scalars stay length 1.

See Also

Other Wind Energy Calculation Functions: barometric_height(), calculate_energy(), get_dist_angles(), turbine_influences()

Crossover Method

Description

Legacy binary crossover. The GA loop uses set_crossover on unique grid-cell IDs instead. This function still permutes 0/1 chromosomes (EQU/RAN) and typically needs trimton afterwards.

Usage

crossover(se6, u, uplimit, crossPart = c("EQU", "RAN"), verbose, seed)

Arguments

Value

Returns a binary coded matrix of all permutations and all grid cells, where 0 indicates no turbine and 1 indicates a turbine in the grid cell.

See Also

Other Genetic Algorithm Functions: fitness(), genetic_algorithm(), init_population(), mutation(), selection(), set_crossover(), swap_mutation(), trimton()

Examples

## Create two random parents with an index and random binary values
Parents <- data.frame(
  ID = 1:20,
  bin = sample(c(0, 1), 20, replace = TRUE, prob = c(70, 30)),
  bin.1 = sample(c(0, 1), 20, replace = TRUE, prob = c(30, 70))
)

## Create random Fitness values for both individuals
FitParents <- data.frame(ID = 1, Fitness = 1000, Fitness.1 = 20)

## Assign both values to a list
CrossSampl <- list(Parents, FitParents)
## Cross their data at equal locations with 2 crossover parts
crossover(CrossSampl, u = 1.1, uplimit = 300, crossPart = "EQU")

## with 3 crossover parts and equal locations
crossover(CrossSampl, u = 2.5, uplimit = 300, crossPart = "EQU")

## or with random locations and 5 crossover parts
crossover(CrossSampl, u = 4.9, uplimit = 300, crossPart = "RAN")

Shiny explorer for a GA result

Description

One-page viewer for a genetic_algorithm() result: Leaflet map of that generation's best layout (downwind wake cones, optional terrain layers), a wind rose, and one plotly figure with fitness, operator rates and population. New fitness maxima are marked; click a marker to jump to that generation. The cell heatmap is not included: use plot_cell_heatmap() or plot_generation() offline. Requires Suggests shiny. Plotly is used when installed.

Usage

explore_result(result, area)

Arguments

Value

The Shiny app object, invisibly. Called for its side effect.

Examples

## Not run: 
explore_result(resultrect, sp_polygon)

## End(Not run)

Evaluate the Individual Fitness values

Description

The fitness of all individuals in the current population is calculated after their energy output has been evaluated in calculate_energy. This function reduces the resulting energy outputs to a single fitness value for each individual.

Usage

fitness(
  population,
  reference_height,
  rotor_height,
  surface_roughness,
  area,
  rotor,
  wind,
  elevation = NULL,
  terrain = FALSE,
  ccl_raster = NULL,
  weibull = FALSE,
  parallel = FALSE,
  n_cluster = 2
)

Arguments

Value

Returns a list with every individual, consisting of X & Y coordinates, rotor radii, the runs and the selected grid cell IDs, and the resulting energy outputs, efficiency rates and fitness values.

See Also

Other Genetic Algorithm Functions: crossover(), genetic_algorithm(), init_population(), mutation(), selection(), set_crossover(), swap_mutation(), trimton()

Examples

## Create a random rectangular shapefile
library(sf)
area <- sf::st_as_sf(sf::st_sfc(
  sf::st_polygon(list(cbind(
    c(4498482, 4498482, 4499991, 4499991, 4498482),
    c(2668272, 2669343, 2669343, 2668272, 2668272)
  ))),
  crs = 3035
))

## Create a uniform and unidirectional wind data.frame and plots the
## resulting wind rose
## Uniform wind speed and single wind direction
wind <- data.frame(ws = 12, wd = 0)
# windrosePlot <- plot_windrose(data = wind, spd = wind$ws,
#                dir = wind$wd, dirres=10, spdmax=20)

## Calculate a Grid and an indexed data.frame with coordinates and
## grid cell IDs.
Grid1 <- grid_area(area = area, size = 200, prop = 1)
Grid <- Grid1[[1]]
AmountGrids <- nrow(Grid)

wind <- list(wind, probab = 100)
startsel <- init_population(Grid, 10, 20)
fit <- fitness(
  population = startsel, reference_height = 100, rotor_height = 100,
  surface_roughness = 0.3, area = area, rotor = 20,
  wind = wind, terrain = FALSE, parallel = FALSE
)

Get or set windfarmGA options

Description

Print every ⁠windfarmGA.*⁠ session option, or set some of them. Names may be given with or without the windfarmGA. prefix.

Usage

ga_options(...)

Arguments

Value

A named list of ⁠windfarmGA.*⁠ options, invisibly.

See Also

options()

Examples

ga_options()
ga_options(immigrants = 3, local_search_tries = 6)

Layouts evaluated in one generation

Description

Return every individual that was fitness-evaluated in a given generation (allCoords). Duplicate layouts (same cell IDs) are flagged.

Usage

generation_layouts(result, generation = NULL)

Arguments

Value

A list with turbines (one row per turbine), layouts (one row per individual) and generation.

See Also

Other Plotting Functions: plot_cell_heatmap(), plot_cloud(), plot_development(), plot_evolution(), plot_generation(), plot_parkfitness(), plot_population(), plot_result(), plot_windfarmGA(), plot_windrose(), population_census(), random_search_single()

Examples

generation_layouts(resultrect, generation = 10)

Run a Genetic Algorithm to optimize a wind farm layout

Description

Run a Genetic Algorithm to optimize the layout of wind turbines on a given area. The algorithm works with a fixed amount of turbines, a fixed rotor radius and a mean wind speed value for every incoming wind direction.

Usage

genetic_algorithm(
  area,
  wind,
  n,
  rotor,
  rotor_height,
  grid_method = "rectangular",
  fcr = 5,
  reference_height = rotor_height,
  surface_roughness = 0.3,
  proportionality = 1,
  iteration = 20,
  mutation_rate = NULL,
  terrain = FALSE,
  elitism = TRUE,
  n_elite = 3,
  selection_mode = "VAR",
  crs = NULL,
  ccl = NULL,
  ccl_roughness = NULL,
  weibull = FALSE,
  weibull_src = NULL,
  parallel = FALSE,
  n_cluster = 2,
  verbose = FALSE,
  plot = FALSE,
  on_generation = NULL
)

Arguments

Details

A terrain effect model can be included in the optimization process. Therefore, a digital elevation model will be downloaded automatically via the elevatr::get_elev_raster function. A land cover raster can also downloaded automatically from the EEA-website, or the path to a raster file can be passed to ccl. The algorithm uses an adapted version of the Raster legend ("clc_legend.csv"), which is stored in the package directory ‘~/inst/extdata’. To use other values for the land cover roughness lengths, insert a column named "Rauhigkeit_z" to the .csv file, assign a surface roughness length to all land cover types. Be sure that all rows are filled with numeric values and save the file with ";" separation. Assign the path of the file to the input variable ccl_roughness of this function.

Fitness is EnergyOverall \times (EfficAllDir/100)^w with w = getOption("windfarmGA.fitness_efficiency_weight"). Hub-height wind speeds use a logarithmic profile unless options(windfarmGA.wind_profile = "power") restores the legacy power law. Power uses options(windfarmGA.Cp) (default 0.45) and optional cut-in / rated / cut-out speeds. Layouts are encoded as n unique grid-cell IDs (set crossover and swap mutation). Selection defaults to VAR (percentage follows fitness progress). Mutation, immigrants and unused-cell injection prefer rarely visited cells. Crossover is spatial with probability options(windfarmGA.spatial_crossover) (default 0.5). Evaluated layouts are cached. A flat global max is not a stop signal. The run ends at iteration, or earlier only after options(windfarmGA.stall_generations) consecutive generations with no new layout, no newly visited cell and no new best fitness (set to 0 to disable). Operator rates cycle like seasons, still only selection / set-crossover / swap-mutation: explore (rates rise on stall) until options(windfarmGA.refine_min_gen) (default 18) and options(windfarmGA.refine_after) (default 12) generations without a new max at coverage \ge 0.35; then refine (inject toward 0.15, mutation toward 2/n, selection toward about 45\ options(windfarmGA.refine_hold) generations in refine (default 25), a short disturbance pulse of options(windfarmGA.explore_pulse) generations (default 10) raises the same rates even if new maxes are still trickling in, then refine resumes. Elites get a short local search each generation: one turbine slides to a neighbouring empty cell (not a random cell anywhere on the grid).

Value

The result is a matrix with aggregated values per generation; the best individual regarding energy and efficiency per generation, some fuzzy control variables per generation, a list of all fitness values per generation, the amount of individuals after each process, a matrix of all energy, efficiency and fitness values per generation, the selection and crossover parameters, a matrix with the generational difference in maximum and mean energy output, a matrix with the given inputs, a dataframe with the wind information, the mutation rate per generation and a matrix with all tested wind farm layouts.

See Also

Other Genetic Algorithm Functions: crossover(), fitness(), init_population(), mutation(), selection(), set_crossover(), swap_mutation(), trimton()

Examples

## Not run: 
## Create a random rectangular shapefile
library(sf)

area <- sf::st_as_sf(sf::st_sfc(
  sf::st_polygon(list(cbind(
    c(4498482, 4498482, 4499991, 4499991, 4498482),
    c(2668272, 2669343, 2669343, 2668272, 2668272)
  ))),
  crs = 3035
))

## Create a uniform and unidirectional wind data.frame and plot the
## resulting wind rose
data.in <- data.frame(ws = 12, wd = 0)
windrosePlot <- plot_windrose(
  data = data.in, spd = data.in$ws,
  dir = data.in$wd, dirres = 10, spdmax = 20
)

## Runs an optimization run for 20 iterations with the
## given shapefile (area), the wind data.frame (data.in),
## 12 turbines (n) with rotor radii of 30m and hub height of 100m.
result <- genetic_algorithm(
  area = area,
  n = 12,
  wind = data.in,
  rotor = 30,
  rotor_height = 100
)
plot_windfarmGA(result = result, area = area)

## End(Not run)

Calculate distances and angles of possibly influencing turbines

Description

Calculate distances and angles for a turbine and all it's potentially influencing turbines.

Usage

get_dist_angles(t, o, wnkl, dist, area, plot_angles = FALSE)

Arguments

Value

Returns a matrix with the distances, angles and heights of potentially influencing turbines

See Also

Other Wind Energy Calculation Functions: barometric_height(), calculate_energy(), circle_intersection(), turbine_influences()

Examples

library(sf)

## Exemplary input Polygon with 2km x 2km:
area <- sf::st_as_sf(sf::st_sfc(
  sf::st_polygon(list(cbind(
    c(4498482, 4498482, 4499991, 4499991, 4498482),
    c(2668272, 2669343, 2669343, 2668272, 2668272)
  ))),
  crs = 3035
))

## Create a random windfarm with 10 turbines
t <- st_coordinates(st_sample(area, 10))
t <- cbind(t, "Z" = 1)
wnkl <- 20
dist <- 100000

## Evaluate and plot for every turbine all other potentially influencing turbines
potInfTur <- list()
for (i in 1:(length(t[, 1]))) {
  potInfTur[[i]] <- get_dist_angles(
    t = t, o = i, wnkl = wnkl,
    dist = dist, area = area, plot_angles = TRUE
  )
}
potInfTur

Map layouts to grid coordinates

Description

Map a population of layouts to grid coordinates. Accepts either an integer matrix of unique cell IDs (n turbines x individuals) or a legacy binary matrix (n_gridcells x individuals).

Usage

get_grids(layouts, grid)

Arguments

Value

Returns a list of all individuals with X and Y coordinates and the grid cell ID.

See Also

Other Helper Functions: grid_area(), hexa_area(), isSpatial(), permutations(), read_power_curve(), splitAt(), wind_from_series(), wind_from_uv(), windata_format()

Examples

## Create a random rectangular shapefile
library(sf)
area <- sf::st_as_sf(sf::st_sfc(
  sf::st_polygon(list(cbind(
    c(0, 0, 2000, 2000, 0),
    c(0, 2000, 2000, 0, 0)
  ))),
  crs = 3035
))

## Calculate a Grid and an indexed data.frame with coordinates and
## grid cell Ids.
Grid1 <- grid_area(area = area, size = 200, prop = 1)
Grid <- Grid1[[1]]

startsel <- init_population(Grid, 10, 20)
wind <- data.frame(ws = 12, wd = 0)
wind <- list(wind, probab = 100)
fit <- fitness(
  population = startsel, reference_height = 100, rotor_height = 100,
  surface_roughness = 0.3, area = area, rotor = 20,
  wind = wind, terrain = FALSE
)
allparks <- do.call("rbind", fit)

## SELECTION (n unique cell IDs per individual)
selec6best <- selection(fit, Grid, 2, TRUE, 6, "VAR")

## Set-crossover and swap-mutation keep exactly n turbines.
cross_ids <- set_crossover(selec6best[[1]], Grid[, "ID"], uplimit = 20)
mut_ids <- swap_mutation(cross_ids, Grid[, "ID"], p = 0.2)

## Look up XY coordinates for the next fitness evaluation.
getRectV <- get_grids(mut_ids, Grid)
fit <- fitness(
  population = getRectV, reference_height = 100, rotor_height = 100,
  surface_roughness = 0.3, area = area, rotor = 20,
  wind = wind, terrain = FALSE
)
head(fit)

Make a grid from a Simple Feature Polygon

Description

Create a grid from a given polygon with a certain resolution and proportionality. The grid cell centroids represent possible wind turbine locations.

Usage

grid_area(area, size = 500, prop = 1, plot_grid = FALSE)

Arguments

Value

Returns a list with 2 elements. List element 1 will have the grid cell IDS, and the X and Y coordinates of the centers of each grid cell. List element 2 is the grid as Simple Feature Polygons, which is used for plotting purposes.

Note

The grid of the genetic algorithm will have a resolution of rotor * fcr. See the arguments of genetic_algorithm().

See Also

Other Helper Functions: get_grids(), hexa_area(), isSpatial(), permutations(), read_power_curve(), splitAt(), wind_from_series(), wind_from_uv(), windata_format()

Examples

## Exemplary input Polygon with 2km x 2km:
library(sf)
area <- sf::st_as_sf(sf::st_sfc(
  sf::st_polygon(list(cbind(
    c(0, 0, 2000, 2000, 0),
    c(0, 2000, 2000, 0, 0)
  ))),
  crs = 3035
))

## Create a Grid
grid_area(area, 200, 1, TRUE)
grid_area(area, 400, 1, TRUE)

## Examplary irregular input Polygon
area <- sf::st_as_sf(sf::st_sfc(
  sf::st_polygon(list(cbind(
    c(0, 0, 2000, 3000, 0),
    c(20, 200, 2000, 0, 20)
  ))),
  crs = 3035
))

## Create a Grid
grid_area(area, 200, 1, TRUE)
grid_area(area, 200, 0.1, TRUE)
grid_area(area, 400, 1, TRUE)
grid_area(area, 400, 0.1, TRUE)

Polygon to Hexagonal Grids

Description

The function takes a Simple Feature Polygon and a size argument and creates a list with an indexed matrix with coordinates and a Simple Feature object, that consists of hexagonal grids.

Usage

hexa_area(area, size = 500, plot_grid = FALSE)

Arguments

Value

Returns a list with 2 elements. List element 1 will have the grid cell IDS, and the X and Y coordinates of the centers of each grid cell. List element 2 is the grid as Simple Feature Polygons, which is used for plotting purposes.

See Also

Other Helper Functions: get_grids(), grid_area(), isSpatial(), permutations(), read_power_curve(), splitAt(), wind_from_series(), wind_from_uv(), windata_format()

Examples

library(sf)
## Exemplary input Polygon with 2km x 2km:
Poly <- sf::st_as_sf(sf::st_sfc(
  sf::st_polygon(list(cbind(
    c(4498482, 4498482, 4499991, 4499991, 4498482),
    c(2668272, 2669343, 2669343, 2668272, 2668272)
  ))),
  crs = 3035
))
HexGrid <- hexa_area(Poly, 100, TRUE)

A POLYGON with a hole

Description

A POLYGON with a hole

Usage

hole_shape

Format

An object of class sf (inherits from data.frame) with 1 rows and 2 columns.

Create a random initial Population

Description

Create n_start random sub-selections from the indexed grid and assign binary variable 1 to selected grids. This function initiates the genetic algorithm with a first random population and will only be needed in the first iteration.

Usage

init_population(grid, n, n_start = 100)

Arguments

Value

Returns a list of n_start initial individuals, each consisting of n turbines. Resulting list has the x and y coordinates, the grid cell ID and a binary variable of 1, indicating a turbine in the grid cell.

See Also

Other Genetic Algorithm Functions: crossover(), fitness(), genetic_algorithm(), mutation(), selection(), set_crossover(), swap_mutation(), trimton()

Examples

library(sf)
## Exemplary input Polygon with 2km x 2km:
area <- sf::st_as_sf(sf::st_sfc(
  sf::st_polygon(list(cbind(
    c(4498482, 4498482, 4499991, 4499991, 4498482),
    c(2668272, 2669343, 2669343, 2668272, 2668272)
  ))),
  crs = 3035
))

Grid <- grid_area(area, 200, 1, TRUE)

## Create 5 individuals with 10 wind turbines each.
firstPop <- init_population(grid = Grid[[1]], n = 10, n_start = 5)

Transform to Simple Feature Polygons

Description

Helper Function, which transforms SpatialPolygons or coordinates in matrix/data.frame - form to a Simple Feature Polygon

Usage

isSpatial(area, crs = NULL)

Arguments

Details

If the columns are named, it will look for common abbreviation to match x/y or long/lat columns. If the columns are not named, the first 2 numeric columns are taken.

Value

A Simple Feature Polygon

See Also

Other Helper Functions: get_grids(), grid_area(), hexa_area(), permutations(), read_power_curve(), splitAt(), wind_from_series(), wind_from_uv(), windata_format()

Examples

library(sf)
df <- rbind(
  c(4498482, 2668272), c(4498482, 2669343),
  c(4499991, 2669343), c(4499991, 2668272)
)
isSpatial(df)

area <- sf::st_as_sf(sf::st_sfc(
  sf::st_polygon(list(cbind(
    c(4498482, 4498482, 4499991, 4499991, 4498482),
    c(2668272, 2669343, 2669343, 2668272, 2668272)
  ))),
  crs = 3035
))
isSpatial(st_coordinates(area), 3035)

A MULTIPOLYGON with 3 Polygons

Description

A MULTIPOLYGON with 3 Polygons

Usage

multi_shape

Format

An object of class sf (inherits from data.frame) with 1 rows and 1 columns.

Mutation Method

Description

Legacy bit-flip mutation on 0/1 chromosomes. The GA loop uses swap_mutation so that every individual keeps exactly n turbines.

Usage

mutation(a, p, seed = NULL)

Arguments

Value

Returns a binary matrix with mutated genes.

See Also

Other Genetic Algorithm Functions: crossover(), fitness(), genetic_algorithm(), init_population(), selection(), set_crossover(), swap_mutation(), trimton()

Examples

## Create 4 random individuals with binary values
a <- cbind(
  bin0 = sample(c(0, 1), 20, replace = TRUE, prob = c(70, 30)),
  bin1 = sample(c(0, 1), 20, replace = TRUE, prob = c(30, 70)),
  bin2 = sample(c(0, 1), 20, replace = TRUE, prob = c(30, 70)),
  bin3 = sample(c(0, 1), 20, replace = TRUE, prob = c(30, 70))
)
a

## Mutate the individuals with a low percentage
aMut <- mutation(a, 0.1, NULL)
## Check which values are not like the originals
a == aMut

## Mutate the individuals with a high percentage
aMut <- mutation(a, 0.4, NULL)
## Check which values are not like the originals
a == aMut

Is the package installed or not

Description

Is the package installed or not

Usage

is_foreach_installed()

is_parallel_installed()

is_doparallel_installed()

is_ggplot2_installed()

is_leaflet_installed()

is_elevatr_installed()

is_plotly_installed()

is_shiny_installed()

Value

An invisible boolean value, indicating if the package is installed or not.

Enumerate the Combinations or Permutations of the Elements of a Vector

Description

permutations enumerates the possible permutations. The function is forked and minified from gtools::permutations

Usage

permutations(n, r, v = 1:n)

Arguments

Value

Returns a matrix where each row contains a vector of length r.

Author(s)

Original versions by Bill Venables. Extended to handle repeats.allowed by Gregory R. Warnes

References

Venables, Bill. "Programmers Note", R-News, Vol 1/1, Jan. 2001. https://cran.r-project.org/doc/Rnews/

See Also

Other Helper Functions: get_grids(), grid_area(), hexa_area(), isSpatial(), read_power_curve(), splitAt(), wind_from_series(), wind_from_uv(), windata_format()

Heatmap of probed grid cells

Description

Count how often each grid cell appeared in an evaluated layout over all generations (allCoords in the GA result). Never-tried cells stay light gray. Useful to see whether the search covered the area or stuck to a few sites.

Usage

plot_cell_heatmap(result, area, log = TRUE, plot = TRUE)

Arguments

Value

A data.frame with grid ID, coordinates and visit count (n_probed), returned invisibly.

See Also

Other Plotting Functions: generation_layouts(), plot_cloud(), plot_development(), plot_evolution(), plot_generation(), plot_parkfitness(), plot_population(), plot_result(), plot_windfarmGA(), plot_windrose(), population_census(), random_search_single()

Examples

plot_cell_heatmap(resultrect, sp_polygon)

Per-generation fitness / efficiency / energy

Description

Summarise every evaluated individual. With pl = TRUE draw three panels of points (max as a line). The returned table has min / mean / max / sd per generation.

Usage

plot_cloud(result, pl = FALSE)

Arguments

Value

A data.frame with fitness, efficiency and energy summaries

See Also

Other Plotting Functions: generation_layouts(), plot_cell_heatmap(), plot_development(), plot_evolution(), plot_generation(), plot_parkfitness(), plot_population(), plot_result(), plot_windfarmGA(), plot_windrose(), population_census(), random_search_single()

Examples

plcdf <- plot_cloud(resulthex, TRUE)

When the best layout improved

Description

Change of the generation-best fitness. Green = new record, orange = flat, red = worse. Blue line is the share of grid cells visited so far.

Usage

plot_development(result)

Arguments

Value

Returns NULL. Used for plotting

See Also

Other Plotting Functions: generation_layouts(), plot_cell_heatmap(), plot_cloud(), plot_evolution(), plot_generation(), plot_parkfitness(), plot_population(), plot_result(), plot_windfarmGA(), plot_windrose(), population_census(), random_search_single()

Examples

plot_development(resultrect)

Energy and efficiency over generations

Description

Max and mean park efficiency and energy yield on one page.

Usage

plot_evolution(result, ask = FALSE, spar = 0.1)

Arguments

Value

Returns NULL. Used for plotting

See Also

Other Plotting Functions: generation_layouts(), plot_cell_heatmap(), plot_cloud(), plot_development(), plot_generation(), plot_parkfitness(), plot_population(), plot_result(), plot_windfarmGA(), plot_windrose(), population_census(), random_search_single()

Examples

plot_evolution(resultrect)

Plot all layouts of one generation

Description

Occupancy map of one generation, then the distinct layouts as small maps. In an interactive session each page waits for Enter so you can step through every distinct layout (n_show maps per page). Non-interactive calls only draw the first n_show maps.

Usage

plot_generation(
  result,
  area,
  generation = NULL,
  n_show = 6,
  interactive = NULL,
  ask = NULL
)

Arguments

Value

The list from generation_layouts, invisibly.

See Also

Other Plotting Functions: generation_layouts(), plot_cell_heatmap(), plot_cloud(), plot_development(), plot_evolution(), plot_parkfitness(), plot_population(), plot_result(), plot_windfarmGA(), plot_windrose(), population_census(), random_search_single()

Examples

plot_generation(resultrect, sp_polygon, generation = 10)

Plot a wind warm with leaflet

Description

Plot a resulting wind farm on a leaflet map. Wakes are downwind cones (Jensen search angle), not circles. Terrain rasters from result$terrainModel are optional overlay layers.

Usage

plot_leaflet(
  result,
  area,
  which = 1,
  orderitems = TRUE,
  grid = NULL,
  wind = NULL,
  terrain = NULL
)

Arguments

Value

Returns a leaflet map.

Examples

## Not run: 
## Plot the best wind farm on a leaflet map (ordered by energy values)
plot_leaflet(result = resulthex, area = sp_polygon, which = 1)

## Plot the last wind farm (ordered by chronology).
plot_leaflet(
  result = resulthex, area = sp_polygon, orderitems = FALSE,
  which = 1
)

## Plot the best wind farm on a leaflet map with the rectangular Grid
Grid <- grid_area(sp_polygon, size = 150, prop = 0.4)
plot_leaflet(
  result = resultrect, area = sp_polygon, which = 1,
  grid = Grid[[2]]
)

## Plot the last wind farm with hexagonal Grid
Grid <- hexa_area(sp_polygon, size = 75)
plot_leaflet(
  result = resulthex, area = sp_polygon, which = 1,
  grid = Grid[[2]]
)

## End(Not run)

Fitness and operator rates

Description

Fitness (max / mean / min) and the three operator rates. Rates are percentages: Selection = share of the population used as parents, Crossover inject = unused cells mixed into children, Mutation = chance that a turbine is swapped to a free cell. The legend sits outside and follows the typical vertical order of the lines (selection highest, then inject, mutation lowest). Plotly hover is used only when ask is FALSE and plotly is installed.

Usage

plot_parkfitness(result, spar = 0.1, interactive = NULL, ask = NULL)

Arguments

Value

A plotly object, or a list of ggplots, invisibly.

See Also

Other Plotting Functions: generation_layouts(), plot_cell_heatmap(), plot_cloud(), plot_development(), plot_evolution(), plot_generation(), plot_population(), plot_result(), plot_windfarmGA(), plot_windrose(), population_census(), random_search_single()

Examples

plot_parkfitness(resulthex)

Plot population size, cells and efficiency

Description

Three pages: (1) individuals / parents / elites, (2) unique grid cells this generation, in the elites, and ever probed, (3) park efficiency. Each page waits for Enter in an interactive session. If the grid has e.g. 70 cells and the whole population still touches every cell, "this generation" and "ever probed" both sit at 70; the elite line is then the one that shrinks toward the best sites.

Usage

plot_population(result, interactive = NULL, ask = NULL)

Arguments

Value

A plotly object, or a list of ggplots, invisibly. The census table is attached as attribute census.

See Also

Other Plotting Functions: generation_layouts(), plot_cell_heatmap(), plot_cloud(), plot_development(), plot_evolution(), plot_generation(), plot_parkfitness(), plot_result(), plot_windfarmGA(), plot_windrose(), population_census(), random_search_single()

Examples

plot_population(resultrect)

Manufacturer power curve

Description

Linear interpolation of a two-column table (ws, power in kW). Set it with ga_options(power_curve = curve) or options(windfarmGA.power_curve = curve). While a curve is set, park energy is the sum of those kW values instead of Cp * v^3. Cut-in / rated / cut-out still apply only when no table is set. If hub wind stays on the rated plateau after wakes, every layout looks the same - use wind in the rising part of the curve. Supply your own table (manufacturer data); the package does not ship copyrighted curves.

Usage

plot_power_curve(curve = NULL, plot = TRUE)

Arguments

Value

The interpolated table, invisibly.

Examples

curve <- data.frame(
  ws = c(0, 3, 4, 8, 12, 25, 26),
  power = c(0, 0, 80, 1200, 2000, 2000, 0)
)
plot_power_curve(curve)

Plot the result of a randomized output.

Description

Plotting method for the results of random_search_single and random_search.

Usage

plot_random_search(resultRS, result, area, best)

Arguments

Value

Returns NULL. Used for plotting

See Also

Other Randomization: random_search(), random_search_single()

Examples

library(sf)
area <- sf::st_as_sf(sf::st_sfc(
  sf::st_polygon(list(cbind(
    c(4498482, 4498482, 4499991, 4499991, 4498482),
    c(2668272, 2669343, 2669343, 2668272, 2668272)
  ))),
  crs = 3035
))

Res <- random_search(result = resultrect, area = area)
plot_random_search(resultRS = Res, result = resultrect, area = area, best = 2)

Plot the best results

Description

Draw the best layout(s) on the site. Turbine labels are the total wake in percent. Default is the single best energy layout.

Usage

plot_result(
  result,
  area,
  best = 1,
  plot_en = 1,
  terrain = FALSE,
  plot_grid = TRUE,
  ccl_roughness = NULL,
  ccl = NULL,
  weibull_src = NULL
)

Arguments

Value

Returns a data.frame of the best (energy/efficiency) individual during all iterations

See Also

Other Plotting Functions: generation_layouts(), plot_cell_heatmap(), plot_cloud(), plot_development(), plot_evolution(), plot_generation(), plot_parkfitness(), plot_population(), plot_windfarmGA(), plot_windrose(), population_census(), random_search_single()

Examples

## Not run: 
## Add some data examples from the package
library(sf)
area <- sf::st_as_sf(sf::st_sfc(
  sf::st_polygon(list(cbind(
    c(4498482, 4498482, 4499991, 4499991, 4498482),
    c(2668272, 2669343, 2669343, 2668272, 2668272)
  ))),
  crs = 3035
))

## Plot the results of a hexagonal grid optimization
plot_result(resulthex, area, best = 1, plot_en = 1, terrain = FALSE)

## Plot the results of a rectangular grid optimization
plot_result(resultrect, area, best = 1, plot_en = 1, terrain = FALSE)

## End(Not run)

Plot visibility

Description

Calculate and plot visibility for given points in a given area. terra::viewshed needs a projected (metre) raster. Lon/lat DEMs (typical elevatr downloads) are projected automatically.

Usage

plot_viewshed(r, turbine_locs, h1 = 0, h2 = 0, plot = TRUE, ...)

Arguments

Value

A mosaiced SpatRaster, representing the visibility for all turbine_locs

Examples

library(sf)
library(terra)

f <- system.file("ex/elev.tif", package = "terra")
r <- rast(f)
x <- project(r, "EPSG:2169")
shape <- sf::st_as_sf(as.polygons(terra::boundaries(x)))
plot(shape)
st_crs(shape) <- 2169
locs <- st_sample(shape, 10, type = "random")
plot_viewshed(x, locs, h1 = 0, h2 = 0, plot = TRUE)

Plot the results of an optimization run

Description

Draw the useful summary plots of a GA run, one after another: best layout, fitness, operator rates, population, cells, efficiency, and the cell heatmap. In an interactive session every page waits for Enter so nothing is overwritten in the Plots pane.

Usage

plot_windfarmGA(
  result,
  area,
  which_plot = "all",
  best = 1,
  plot_en = 1,
  weibull_src = NULL,
  ask = NULL,
  plotly = NULL
)

Arguments

Value

Returns NULL. Used for plotting

See Also

Other Plotting Functions: generation_layouts(), plot_cell_heatmap(), plot_cloud(), plot_development(), plot_evolution(), plot_generation(), plot_parkfitness(), plot_population(), plot_result(), plot_windrose(), population_census(), random_search_single()

Examples

## Not run: 
library(sf)
area <- sf::st_as_sf(sf::st_sfc(
  sf::st_polygon(list(cbind(
    c(4498482, 4498482, 4499991, 4499991, 4498482),
    c(2668272, 2669343, 2669343, 2668272, 2668272)
  ))),
  crs = 3035
))

plot_windfarmGA(resulthex, area)
plot_windfarmGA(resultrect, area, which_plot = "progress")

## End(Not run)

Plot a Windrose

Description

Plot a wind rose of the wind data frame.

Usage

plot_windrose(
  data,
  spd,
  dir,
  spdres = 2,
  dirres = 10,
  spdmin = 1,
  spdmax = 30,
  palette = "YlGnBu",
  spdseq = NULL,
  plot = TRUE
)

Arguments

Value

A ggplot2 wind rose plot, returned invisibly.

See Also

Other Plotting Functions: generation_layouts(), plot_cell_heatmap(), plot_cloud(), plot_development(), plot_evolution(), plot_generation(), plot_parkfitness(), plot_population(), plot_result(), plot_windfarmGA(), population_census(), random_search_single()

Examples

## Exemplary Input Wind speed and direction data frame
# Uniform wind speed and single wind direction
data.in <- data.frame(ws = 12, wd = 0)
windrosePlot <- plot_windrose(
  data = data.in, spd = data.in$ws,
  dir = data.in$wd
)

# Random wind speeds and random wind directions
data.in <- data.frame(
  ws = sample(1:25, 10),
  wd = sample(1:260, 10)
)
windrosePlot <- plot_windrose(
  data = data.in, spd = data.in$ws,
  dir = data.in$wd
)

Population size and diversity per generation

Description

Counts evaluated individuals, distinct layouts, duplicates dropped before the next generation, selected parents, elites, elite offspring and grid cells (this generation and cumulative).

Usage

population_census(result)

Arguments

Value

A data.frame with one row per generation.

See Also

Other Plotting Functions: generation_layouts(), plot_cell_heatmap(), plot_cloud(), plot_development(), plot_evolution(), plot_generation(), plot_parkfitness(), plot_population(), plot_result(), plot_windfarmGA(), plot_windrose(), random_search_single()

Examples

population_census(resultrect)

Randomize the output of the Genetic Algorithm

Description

Jitter the best GA layouts inside their grid cells and re-evaluate energy. Use this as a short post-search after genetic_algorithm(). Terrain and Weibull follow the GA flags when terrain / weibull are NULL. Terrain rasters from the GA are reused when stored in result; pass a DEM to rebuild. Weibull rasters are not stored; pass weibull_src again if needed.

Usage

random_search(
  result,
  area,
  runs = 20,
  best = 1,
  plot = FALSE,
  max_dist = 2.2,
  terrain = NULL,
  weibull = NULL,
  weibull_src = NULL,
  ccl = NULL,
  ccl_roughness = NULL
)

Arguments

Value

Returns a list.

See Also

Other Randomization: plot_random_search(), random_search_single()

Examples

new <- random_search(resultrect, sp_polygon, runs = 20, best = 4)
plot_random_search(resultRS = new, result = resultrect, area = sp_polygon, best = 2)

Randomize the location of a single turbine

Description

Perform a random search for a single turbine, to further optimize the output of the wind farm layout.

Usage

random_search_single(
  result,
  area,
  runs = 20,
  plot = FALSE,
  max_dist = 2.2,
  terrain = NULL,
  weibull = NULL,
  weibull_src = NULL,
  ccl = NULL,
  ccl_roughness = NULL,
  turbine = NULL
)

Arguments

Value

Returns a list

See Also

Other Randomization: plot_random_search(), random_search()

Other Plotting Functions: generation_layouts(), plot_cell_heatmap(), plot_cloud(), plot_development(), plot_evolution(), plot_generation(), plot_parkfitness(), plot_population(), plot_result(), plot_windfarmGA(), plot_windrose(), population_census()

Read a manufacturer or NREL/IEA power-curve table

Description

Parse a CSV (or data.frame) with wind speed and power in kW. Understands NREL/IEA headers such as ⁠Wind Speed [m/s]⁠ and Power [kW]. Optional Ct is returned as an attribute. Does not ship manufacturer curves; pass your own file or an open IEA/NREL reference CSV. See experimental/climate_helpers.R.

Usage

read_power_curve(file)

Arguments

Value

A data.frame with ws and power. Attribute ct is a matching data.frame when a thrust column is present.

See Also

Other Helper Functions: get_grids(), grid_area(), hexa_area(), isSpatial(), permutations(), splitAt(), wind_from_series(), wind_from_uv(), windata_format()

Examples

curve <- data.frame(
  `Wind Speed [m/s]` = c(3, 8, 12, 25),
  `Power [kW]` = c(0, 1200, 2000, 2000),
  check.names = FALSE
)
read_power_curve(curve)

A resulting matrix of genetic_algorithm with 10 iterations and a hexagonal grid derived from sp_polygon

Description

A resulting matrix of genetic_algorithm with 10 iterations and a hexagonal grid derived from sp_polygon

Usage

resulthex

Format

An object of class matrix (inherits from array) with 10 rows and 13 columns.

A resulting matrix of genetic_algorithm with 50 generations and a rectangular grid derived from sp_polygon

Description

First 50 generations of a 200-iteration run, kept short so the installed package stays small. Do not recreate with the current GA (n_start and population size are larger; the .rda grows).

Usage

resultrect

Format

An object of class matrix (inherits from array) with 50 rows and 13 columns.

Selection Method

Description

Select a certain amount of individuals and recombine them to parental teams. Add the mean fitness value of both parents to the parental team. Depending on the selected selection_mode, the algorithm will either take always 50 percent or a variable percentage of the current population. The variable percentage depends on the evolution of the populations fitness values. With elitism = TRUE the best individuals are always included in the mating pool.

Usage

selection(
  fit,
  grid,
  share,
  elitism = TRUE,
  n_elite = 3,
  selection_mode = "VAR",
  verbose = FALSE
)

Arguments

Value

Returns a list with 2 elements. Element 1 is an integer matrix of selected layouts (n turbines x selected individuals), each column a set of unique grid cell IDs. Element 2 is the fitness of each selected individual.

See Also

Other Genetic Algorithm Functions: crossover(), fitness(), genetic_algorithm(), init_population(), mutation(), set_crossover(), swap_mutation(), trimton()

Examples

## Exemplary input Polygon with 2km x 2km:
library(sf)
area <- sf::st_as_sf(sf::st_sfc(
  sf::st_polygon(list(cbind(
    c(4498482, 4498482, 4499991, 4499991, 4498482),
    c(2668272, 2669343, 2669343, 2668272, 2668272)
  ))),
  crs = 3035
))

## Calculate a Grid and an indexed data.frame with coordinates and grid cell Ids.
Grid1 <- grid_area(area = area, size = 200, prop = 1)
Grid <- Grid1[[1]]
AmountGrids <- nrow(Grid)

startsel <- init_population(Grid, 10, 20)
wind <- as.data.frame(cbind(ws = 12, wd = 0))
wind <- list(wind, probab = 100)
fit <- fitness(
  population = startsel, reference_height = 100, rotor_height = 100,
  surface_roughness = 0.3, area = area, rotor = 20, wind = wind,
  terrain = FALSE
)
allparks <- do.call("rbind", fit)
## SELECTION
## print the amount of Individuals selected. Check if the amount
## of Turbines is as requested.
selec6best <- selection(fit, Grid, 2, TRUE, 6, "VAR")
selec6best <- selection(fit, Grid, 2, TRUE, 6, "FIX")
selec6best <- selection(fit, Grid, 4, FALSE, 6, "FIX")

Set crossover of turbine layouts

Description

Combine two layouts of n unique grid-cell IDs. Shared sites are kept; remaining sites are sampled from the parents' exclusive cells and, with rate p_inject, from grid cells that neither parent uses. Identical parents still get unused cells injected so the search does not freeze. Every child has exactly n turbines.

Usage

set_crossover(
  ids,
  grid_ids,
  uplimit = 300,
  seed = NULL,
  verbose = FALSE,
  p_inject = NULL,
  grid_xy = NULL,
  visit = NULL,
  p_spatial = NULL
)

Arguments

Value

Integer matrix of unique grid IDs (n x children)

See Also

Other Genetic Algorithm Functions: crossover(), fitness(), genetic_algorithm(), init_population(), mutation(), selection(), swap_mutation(), trimton()

Examples

ids <- cbind(c(1, 3, 5, 7), c(1, 4, 5, 9))
set_crossover(ids, grid_ids = 1:20, uplimit = 4, seed = 1)

The rectangular POLYGON used to create resultrect & resulthex

Description

The rectangular POLYGON used to create resultrect & resulthex

Usage

sp_polygon

Format

An object of class sf (inherits from data.frame) with 1 rows and 1 columns.

Split matrices or numeric vectors at specific indices

Description

The function is used by the crossover method to split a genetic code at certain intervals. See also crossover.

Usage

splitAt(x, pos)

Arguments

Value

Returns a list of the split genetic code.

See Also

Other Helper Functions: get_grids(), grid_area(), hexa_area(), isSpatial(), permutations(), read_power_curve(), wind_from_series(), wind_from_uv(), windata_format()

Examples

splitAt(1:100, 20)
splitAt(as.matrix(1:100), 20)

Swap mutation of turbine layouts

Description

Replace occupied grid cells with unused ones. The number of swaps is max(min_swaps, Binomial(n, p)), so each individual explores at least min_swaps new cells (default 1). The number of turbines stays n.

Usage

swap_mutation(ids, grid_ids, p, seed = NULL, min_swaps = NULL, visit = NULL)

Arguments

Value

Integer matrix of unique grid IDs, same dimension as ids

See Also

Other Genetic Algorithm Functions: crossover(), fitness(), genetic_algorithm(), init_population(), mutation(), selection(), set_crossover(), trimton()

Examples

ids <- cbind(c(1, 3, 5, 7), c(2, 4, 6, 8))
swap_mutation(ids, grid_ids = 1:20, p = 0.5, seed = 1)

Get terrainhic rasters

Description

Calculate the SpatRasters needed for the terrain model.

Usage

terrain_model(
  terrain = TRUE,
  area,
  ccl,
  ccl_roughness,
  plot = FALSE,
  verbose = FALSE
)

Arguments

Value

A list of SpatRasters

Examples

## Not run: 
library(sf)
area <- sf::st_as_sf(sf::st_sfc(
  sf::st_polygon(list(cbind(
    c(4651704, 4651704, 4654475, 4654475, 4651704),
    c(2692925, 2694746, 2694746, 2692925, 2692925)
  ))),
  crs = 3035
))
Polygon_wgs84 <- sf::st_transform(area, st_crs(4326))
srtm <- elevatr::get_elev_raster(locations = Polygon_wgs84, z = 11)
res <- terrain_model(srtm, area)

## End(Not run)

Adjust the amount of turbines per windfarm

Description

Legacy repair for binary chromosomes. The GA loop encodes layouts as n unique grid IDs, so this function is not called there. It remains exported for the old 0/1 pipeline (crossover / mutation).

Usage

trimton(mut, nturb, allparks, nGrids, trimForce, seed)

Arguments

Value

Returns a binary matrix with the correct amount of turbines per individual

See Also

Other Genetic Algorithm Functions: crossover(), fitness(), genetic_algorithm(), init_population(), mutation(), selection(), set_crossover(), swap_mutation()

Examples

## Create a random rectangular shapefile
library(sf)
area <- sf::st_as_sf(sf::st_sfc(
  sf::st_polygon(list(cbind(
    c(0, 0, 2000, 2000, 0),
    c(0, 2000, 2000, 0, 0)
  ))),
  crs = 3035
))

## Create a uniform and unidirectional wind data.frame and plots the
## resulting wind rose
## Uniform wind speed and single wind direction
data.in <- as.data.frame(cbind(ws = 12, wd = 0))

## Calculate a Grid and an indexed data.frame with coordinates and grid cell Ids.
Grid1 <- grid_area(area = area, size = 200, prop = 1)
Grid <- Grid1[[1]]
AmountGrids <- nrow(Grid)

startsel <- init_population(Grid, 10, 20)
wind <- as.data.frame(cbind(ws = 12, wd = 0))
wind <- list(wind, probab = 100)
fit <- fitness(
  population = startsel, reference_height = 100, rotor_height = 100,
  surface_roughness = 0.3, area = area, rotor = 20,
  wind = wind, terrain = FALSE
)
allparks <- do.call("rbind", fit)
## selection() returns ID matrices; crossover()/trimton() expect 0/1.
sel <- selection(fit, Grid, 2, TRUE, 6, "FIX")
ids <- sel[[1]]
bins <- matrix(0, nrow(Grid), ncol(ids))
for (j in seq_len(ncol(ids))) {
  bins[match(ids[, j], Grid[, "ID"]), j] <- 1
}
selec6best <- list(
  data.frame(ID = Grid[, "ID"], bins),
  data.frame(ID = 1, t(sel[[2]]))
)
crossOut <- crossover(selec6best, 2, uplimit = 300, crossPart = "RAN")
mut <- mutation(a = crossOut, p = 0.3, NULL)
mut1 <- trimton(
  mut = mut, nturb = 10, allparks = allparks, nGrids = AmountGrids,
  trimForce = FALSE
)
colSums(mut)
colSums(mut1)

Find potentially influencing turbines

Description

Find all turbines that could potentially influence another turbine and save them to a list.

Usage

turbine_influences(t, wnkl, dist, area, dirct, plot_angles = FALSE)

Arguments

Value

Returns a list of all individuals of the current generation which could potentially influence other turbines. List includes the relevant coordinates, the distances and angles in between and assigns the Point ID.

See Also

Other Wind Energy Calculation Functions: barometric_height(), calculate_energy(), circle_intersection(), get_dist_angles()

Examples

## Exemplary input Polygon with 2km x 2km:
library(sf)

area <- sf::st_as_sf(sf::st_sfc(
  sf::st_polygon(list(cbind(
    c(0, 0, 2000, 2000, 0),
    c(0, 2000, 2000, 0, 0)
  ))),
  crs = 3035
))

t <- st_coordinates(st_sample(area, 10))
t <- cbind(t, "Z" = 1)
wnkl <- 20
dist <- 100000
dirct <- 0

res <- turbine_influences(t, wnkl, dist, area, dirct, plot_angles = TRUE)

Downwind wake search cones

Description

Build sf polygons for the Jensen search cone of each turbine and wind direction. Colour can encode wake loss (AbschGesamt).

Usage

wake_cones(xy, wind, rotor, area, half_deg = NULL, colors = NULL)

Arguments

Value

An sf polygon layer with turb, wd, prob, farbe.

Wind rose from speed and direction series

Description

Bin a time series of hub-height speed and meteorological direction into ws / wd / probab.

Usage

wind_from_series(ws, wd, dir_width = 30)

Arguments

Value

A data.frame with ws, wd, probab.

See Also

Other Helper Functions: get_grids(), grid_area(), hexa_area(), isSpatial(), permutations(), read_power_curve(), splitAt(), wind_from_uv(), windata_format()

Wind rose from u/v components

Description

Bin ERA5-style eastward (u) and northward (v) wind components into the ws / wd / probab table that windata_format() and genetic_algorithm() expect. Direction is meteorological (where the wind comes from; 0 = north). Each direction bin gets the mean speed and the hour share.

Usage

wind_from_uv(u, v, dir_width = 30)

Arguments

Value

A data.frame with ws, wd, probab (probabilities sum to 100).

See Also

Other Helper Functions: get_grids(), grid_area(), hexa_area(), isSpatial(), permutations(), read_power_curve(), splitAt(), wind_from_series(), windata_format()

Examples

set.seed(1)
u <- rnorm(200, 2)
v <- rnorm(200, -3)
wind_from_uv(u, v, dir_width = 30)

Transform Winddata

Description

Helper Function, which transforms winddata to an acceptable format

Usage

windata_format(df)

Arguments

Value

A list of windspeed and probabilities

See Also

Other Helper Functions: get_grids(), grid_area(), hexa_area(), isSpatial(), permutations(), read_power_curve(), splitAt(), wind_from_series(), wind_from_uv()

Examples

wind_df <- data.frame(
  ws = c(12, 30, 45),
  wd = c(0, 90, 150),
  probab = 30:32
)
windata_format(wind_df)

wind_df <- data.frame(
  speed = c(12, 30, 45),
  direction = c(90, 90, 150),
  probab = c(10, 20, 60)
)
windata_format(wind_df)

wind_df <- data.frame(
  speed = c(12, 30, 45),
  direction = c(400, 90, 150)
)
windata_format(wind_df)