windfarmGA (version 2.3.0)

selection: 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 selstate, 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.

Usage

selection(fit, Grid, teil, elitism, nelit, selstate, verbose)

Arguments

fit

A list of all fitness-evaluated individuals

Grid

Is the indexed grid output from grid_area

teil

A numeric value that determines the selection percentage

elitism

Boolean value which indicates whether elitism should be included or not.

nelit

If elitism is TRUE, then this input variable determines the amount of individuals in the elite group.

selstate

Determines which selection method is used, "FIX" selects a constant percentage and "VAR" selects a variable percentage, depending on the development of the fitness values.

verbose

If TRUE, will print out further information.

Value

Returns list with 2 elements. Element 1 is the binary encoded matrix which shows all selected individuals. Element 2 represent the mean fitness values of each parental team.

See Also

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

Examples

# NOT RUN {
## Create a random rectangular shapefile
library(sp)
Polygon1 <- Polygon(rbind(c(0, 0), c(0, 2000), c(2000, 2000), c(2000, 0)))
Polygon1 <- Polygons(list(Polygon1),1);
Polygon1 <- SpatialPolygons(list(Polygon1))
Projection <- "+init=epsg:3035"
proj4string(Polygon1) <- CRS(Projection)

## Calculate a Grid and an indexed data.frame with coordinates and grid cell Ids.
Grid1 <- grid_area(shape = Polygon1,resol = 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(selection = startsel, referenceHeight = 100, RotorHeight=100,
               SurfaceRoughness=0.3,Polygon = Polygon1, resol1 = 200,
               rot = 20, dirspeed = wind, 
               srtm_crop = "", topograp = FALSE, cclRaster = "")
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")
# }