Abstract: The Markov decision process (MDP) based
methodology is implemented in order to establish the optimal
schedule which minimizes the cost. Formulation of MDP problem
is presented using the information about the current state of pipe,
improvement cost, failure cost and pipe deterioration model. The
objective function and detailed algorithm of dynamic programming
(DP) are modified due to the difficulty of implementing the
conventional DP approaches. The optimal schedule derived from
suggested model is compared to several policies via Monte
Carlo simulation. Validity of the solution and improvement in
computational time are proved.
Abstract: This paper presents a new problem solving approach
that is able to generate optimal policy solution for finite-state
stochastic sequential decision-making problems with high data
efficiency. The proposed algorithm iteratively builds and improves
an approximate Markov Decision Process (MDP) model along with
cost-to-go value approximates by generating finite length trajectories
through the state-space. The approach creates a synergy between an
approximate evolving model and approximate cost-to-go values to
produce a sequence of improving policies finally converging to the
optimal policy through an intelligent and structured search of the
policy space. The approach modifies the policy update step of the
policy iteration so as to result in a speedy and stable convergence to
the optimal policy. We apply the algorithm to a non-holonomic
mobile robot control problem and compare its performance with
other Reinforcement Learning (RL) approaches, e.g., a) Q-learning,
b) Watkins Q(λ), c) SARSA(λ).