Generate TOPIA weather and demand profiles from a closed-form model of the
map, at any resolution and for any cluster roster. The shipped "simulated"
dataset (topia) is one cached run of this function, so
topia_simulate() at its defaults reproduces it and every other parameter
set is an alternative.
The output is shaped like topia_profiles()'s, so either can be passed to
build_topia(weather = , demand = ).
Usage
topia_simulate(
regions = NULL,
calendar = "s4_h24",
clusters = NULL,
solar_season = .SIM_M_SOL,
wind_season = .SIM_M_WIN,
hydro_season_calm = .SIM_M_HYD_A,
hydro_season_windy = .SIM_M_HYD_B,
wind_day_calm = .SIM_H_WIN_CALM,
wind_day_windy = .SIM_H_WIN_WINDY,
load_day_res = .SIM_H_RES,
load_day_ind = .SIM_H_IND,
load_season_res = .SIM_M_RES,
load_season_ind = .SIM_M_IND,
residential = .SIM_W_RES,
wind_grade = .SIM_WIN_GRADE,
solar_halflen = .SIM_SOL_HALFLEN,
hydro_cf = .SIM_HYD_CF,
jitter = 0,
seed = NULL,
layout = "honeycomb",
measured = NULL
)Arguments
- regions
provinces to generate, defaulting to all of TOPIA's.
- calendar
a calendar name, a calendar object, or a timetable
data.frame. It must carry anHOURframe and eitherYDAYorSEASONabove it.- clusters
the roster to generate: a
data.frameofresource,clusterandregion, orNULLto take the rosters the shipped weather objects declare.FALSEgenerates one ungraded series per resource.- solar_season, wind_season, hydro_season_calm, hydro_season_windy
twelve monthly multipliers each, normalised to mean 1 internally.
- wind_day_calm, wind_day_windy, load_day_res, load_day_ind
twenty-four hourly values each: the endpoints the shapes are blended between.
- load_season_res, load_season_ind
twelve monthly multipliers each.
- residential
residential share of load by zone, named.
- wind_grade, solar_halflen
the cluster shape parameters, named by cluster.
- hydro_cf
annual mean inflow factor, used where the measured set has no hydro series for a province.
- jitter
relative amplitude of a deterministic perturbation applied to each field,
0for none. The mean is preserved and no values are made equal, so the energy is unbiased and the LP stays non-degenerate.- seed
integer seed for
jitter.- layout
the geoscale whose zones the load mix is keyed on.
- measured
the measured frames the levels are calibrated against, as
list(weather =, demand =).NULLreads the shipped set, which is what a user wants; the dataset build passes its own frames, since at that point the shipped set is what is being written.
Value
a list of two tidy data.frames: weather
(resource, cluster, region, timeslice, wval) and demand
(region, timeslice, load), both carrying a dataset attribute
naming the generator and the parameters that differ from the defaults.
Details
Levels are calibrated against the shipped measured set: each
(resource, cluster, province) field is scaled to the mean capacity factor
measured there, and the load shape to the measured peak-to-mean. What the
arguments change is the SHAPE.
Clusters differ through one parameter each – wind_grade offsets a grade
along the calm-to-windy axis, solar_halflen sets the half-length of the
solar day per mounting – so a grade is a different resource rather than
another ceiling on one series. A roster with no clusters yields one series
per resource with cluster NA, which no graded technology can use; see
build_topia(clusters = ).
See also
topia_profiles() for the shipped sets, build_topia(),
topia_profile() for geometric shapes
Other topia:
topia,
topia_profile(),
topia_profiles()
Examples
p <- topia_simulate(regions = c("R01", "R10"), calendar = "s4_h24")
head(p$weather, 3)
#> resource cluster region timeslice wval
#> 1 WSOL FL R01 WIN_h00 0
#> 2 WSOL FL R01 WIN_h01 0
#> 3 WSOL FL R01 WIN_h02 0
# an alternative: no seasonality in the wind
p2 <- topia_simulate(regions = c("R01", "R10"), wind_season = rep(1, 12))
