MCVI.MCVISolver
— TypeMCVISolver <: POMDPs.Solver
The MCVI solver is a solver for POMDPs that uses Monte Carlo Value Iteration to solve the problem. Described in Bai, H., Hsu, D., & Lee, W. S. (2014). Integrated perception and planning in the continuous space: A POMDP approach. The International Journal of Robotics Research, 33(9), 1288-1302.
Fields
simulater::POMDPs.Simulator
root::Union{BeliefNode, Nothing}
n_iter::Int64
: Number of iterationsnum_particles::Int64
: Number of belief particles to be usedobs_branch::Int64
: Branching factor (previous default: 8)num_state::Int64
: Number of states to sample from belief (previous default: 500)num_prune_obs::Int64
: Number of times to sample observation while pruning alpha edges (previous default: 1000)num_eval_belief::Int64
: Number of times to simulate while evaluating belief (previous default: 5000)num_obs::Int64
: Number of observations to sample while evaluating belief (previous default: 50)lbound::Any
: An object representing the lower bound. The functionMCVI.lower_bound(lbound, problem, s)
will be called to get the lower bound for the states
- this function needs to be implemented for the solver to work.ubound::Any
: An object representing the upper bound. The functionMCVI.upper_bound(ubound, problem, s)
will be called to get the lower bound for the states
- this function needs to be implemented for the solver to work.
Reference the docs for an example of bounds implemented for the Light Dark problem.