Calculate water values with Grid_Matrix from estimated reward, used in calculateBellmanWithIterativeSimulations

Calculate water values with Grid_Matrix from estimated reward, used in calculateBellmanWithIterativeSimulations

Usage

updateWatervalues(
  reward,
  controls,
  area,
  mcyears,
  opts,
  states_step_ratio,
  pump_eff,
  penalty_low,
  penalty_high,
  inflow,
  niveau_max,
  max_hydro_weekly,
  cvar_value = 1,
  final_level,
  penalty_final_level = NULL
)

Arguments

reward

Data frame containing estimation of the reward function, same format as the output of reward_offset, not yet offseted with respect to 0

controls

Data frame containing possible transition for each week, generated by the function constraint_generator

area

Area with the reservoir

mcyears

Vector of monte carlo years used to evaluate rewards

opts

List of simulation parameters returned by the function antaresRead::setSimulationPath

states_step_ratio

Discretization ratio to generate steps levels between the reservoir capacity and zero

pump_eff

Pumping efficiency between 0 and 1 (1 if no pumping)

penalty_low

Penalty for violating the bottom rule curve

penalty_high

Penalty for violating the top rule curve

inflow

Data frame with inflows for each week and each scenario, generated by the function antaresRead::readInputTS

niveau_max

Capacity of the reservoir in MWh

max_hydro_weekly

data.frame timeId,pump,turb with maximum pumping and storing powers for each week,returned by the function get_max_hydro

cvar_value

from 0 to 1. the probability used in quantile method to determine a bellman value which cvar_value all bellman values are equal or less to it. (quantile(cvar_value))

final_level

Final level (in percent between 0 and 100) if final level is constrained but different from initial level

penalty_final_level

Penalties (for both bottom and top rule curves) to constrain final level

Value

List containing aggregated water values and the data table with all years