Calculate Bellman values sequentially, one area at a time.

For each area, reward functions are first computed using calculateRewardsSimulationsWithPlaia(). Bellman values are then computed with Grid_Matrix(), and an optimal trajectory for the area is derived using getOptimalTrend(). The resulting trajectory is subsequently used to compute reward functions for the next area in the sequence. For areas where Bellman values have not yet been computed, either initial_traj or a short-term trajectory is used as a fallback.

Usage

getBellmanValuesSequentialMultiStockWithPlaia(
  list_areas,
  opts,
  nb_simulations,
  mcyears,
  penalty_low,
  penalty_high,
  states_step_ratio = 1/50,
  cvar_value = 1,
  penalty_final_level = NULL,
  initial_traj = NULL,
  list_areas_to_compute = NULL
)

Arguments

list_areas

Vector of areas concerned by simulations.

opts

List of study parameters returned by the function antaresRead::setSimulationPath(simulation="input") in input mode.

nb_simulations

Number of controls to simulate

mcyears

Vector of integer. Monte Carlo years used to compute water values.

penalty_low

Double. Penalty for violating the bottom rule curve, comparable to the unsupplied energy cost.

penalty_high

Double. Penalty for violating the top rule curve, comparable to the spilled energy cost.

states_step_ratio

Double. Discretization ratio to generate steps levels between the reservoir capacity and zero for which Bellman values are computed.

cvar_value

Double from 0 to 1. The probability used in cvar method.

penalty_final_level

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

initial_traj

Initial trajectory (used for other storages)

list_areas_to_compute

Vector of character. Areas for which to compute Bellman values. If NULL, all areas in list_areas are used.

Value

List containing aggregated water values, reward functions and optimal trajectories.