DeepAI AI Chat
Log In Sign Up

Analytical and numerical solutions to ergodic control problems arising in environmental management

by   Hidekazu Yoshioka, et al.

Environmental management should be based on a long-run sustainable viewpoint. A stochastic control problem optimizing a long-run objective function is an ergodic control problem whose resolution can be achieved by solving an associated Hamilton-Jacobi-Bellman (HJB) equation having an effective Hamiltonian. However, this key topic has not been well-studied in the context of environmental management. Focusing on non-smooth sediment storage management as an engineering problem, we formulate a new ergodic control problem under discrete observations: a natural assumption of realistic observations. We give an analytical solution to a simple problem and verify optimality and uniqueness of the corresponding HJB equation. To manage HJB equations in more general cases, we propose a fast sweep method resorting to neither pseudo-time integration nor vanishing discount. The optimal policy and the effective Hamiltonian are then computed simultaneously. A robust control counterpart where the dynamics involve uncertainties is also considered based on a multiplier robust formalism that harmonizes with the stochastic control approach. The resulting HJB equation has a stronger non-linearity but is managed using the presented numerical method. An extended problem constraining control policy is analyzed as well.


A random observation-based management model of population dynamics and its ecological application

A new stochastic control problem of population dynamics under partial ob...

Regime-switching constrained viscosity solutions approach for controlling dam-reservoir systems

A new stochastic control problem of a dam-reservoir system installed in ...

A control method for solving high-dimensional Hamiltonian systems through deep neural networks

In this paper, we mainly focus on solving high-dimensional stochastic Ha...

Ternary Policy Iteration Algorithm for Nonlinear Robust Control

The uncertainties in plant dynamics remain a challenge for nonlinear con...