Abstract:The paper is devoted to the optimal control of a system with two time-scales, in a regime when the limit equation is not of averaging type but, in the spirit of Wong-Zakai principle, it is a stochastic differential equation for the slow variable, with noise emerging from the fast one. It proves that it is possible to control the slow variable by acting only on the fast scales. The concrete problem, of interest for climate research, is embedded into an abstract framework in Hilbert spaces, with a stochastic process driven by an approximation of a given noise. The principle established here is that convergence of the uncontrolled problem is sufficient for convergence of both the optimal costs and the optimal controls. This target is reached using Girsanov transform and the representation of the optimal cost and the optimal controls using a Forward Backward System. A challenge in this program is represented by the generality considered here of unbounded control actions.

What problem does this paper attempt to address?

The main problem that this paper attempts to solve is whether in a stochastic system with two - time scales, when the limit equation is not of the averaging type but a stochastic differential equation of slow variables (with noise coming from fast variables), the slow variables can be effectively controlled by only controlling the fast variables. Specifically, the focus of the research is to explore specific problems of great significance in climate research and embed them into an abstract framework, using stochastic processes in Hilbert space and their noise approximations. ### Detailed description of the main problem 1. **Optimal control of two - time - scale systems**: - The research object is a system with two - time scales. In a certain limit case, the limit equation of the slow variable is not a simple averaging type but a stochastic differential equation with noise generated by the fast variable. - The goal is to prove that by only controlling the fast variable, the behavior of the slow variable can be effectively controlled. 2. **Convergence problem**: - The paper explores whether the convergence of the uncontrolled problem is sufficient to guarantee the convergence of the optimal cost and the optimal control when the parameter \(\epsilon \to 0\). - Specifically, it is assumed that if the solution of the uncontrolled equation converges in probability, then the optimal cost and the optimal control should also converge. 3. **Application in climate models**: - The paper applies the above - mentioned theory to climate models, especially slow - fast systems, where the slow variable represents changes on the climate scale and the fast variable represents changes on the meteorological scale. - The key question is: Can the evolution of the climate be "controlled" by only applying control on the meteorological scale? ### Mathematical models and methods - **Approximation of stochastic optimal control problems**: - A family of stochastic equations specified by the parameter \(\epsilon \in (0,1)\) is introduced, which are driven by approximations of a given noise \(W\). - For each \(\epsilon\), a control problem is solved, and the convergence of the optimal cost and the optimal control when \(\epsilon \to 0\) is studied. - **Forward - backward stochastic differential equations (FBSDE)**: - FBSDE is used to represent the optimal cost and the optimal control, which helps to characterize the optimal state and the optimal feedback law without the regularity of the value function. - **BMO martingale techniques**: - The properties and estimates of BMO martingales are utilized to ensure the well - posedness and convergence of the optimal control problem within the weak - form framework. ### Main results - **Convergence theorem**: - It is proved that the convergence of the uncontrolled problem is sufficient to guarantee the convergence of the optimal cost and the optimal control (see Theorem 4.2). - This means that by only applying control on the fast variable, the long - term behavior of the slow variable can be effectively controlled when \(\epsilon \to 0\). ### Application background - **Significance in climate research**: - In the context of climate - weather interactions, the research results show that by applying appropriate control on the meteorological scale, the long - term evolution of the climate can be influenced. - This provides a theoretical basis for understanding how to regulate long - term climate change through short - term meteorological interventions. In conclusion, through rigorous mathematical analysis, this paper explores the possibility of influencing slow variables by controlling fast variables in complex stochastic systems and applies it to climate models, providing a new theoretical basis for climate regulation.

On approximations of stochastic optimal control problems with an application to climate equations

Stochastic Optimal Control of Quasi Non-Integrable Hamiltonian Systems with Stochastic Maximum Principle

Stochastic Optimal Control for the Response of Quasi Non-Integrable Hamiltonian Systems

STOCHASTIC OPTIMAL CONTROL OF UNCERTAIN QUASI-HAMILTONIAN SYSTEMS

Brief paper: Stochastic minimax control for stabilizing uncertain quasi- integrable Hamiltonian systems

On Stochastic Optimal Control of Partially Observable Nonlinear Quasi Hamiltonian Systems

A Multilevel Approach for Stochastic Nonlinear Optimal Control

Optimal control of diffusion processes: $\infty$-order variational analysis and numerical solution

Optimal control for two-dimensional stochastic second grade fluids

Particle Methods For Stochastic Optimal Control Problems

Finite element approximations for stochastic control problems with unbounded state space

Discrete-Time Approximation of Stochastic Optimal Control with Partial Observation

Optimal Control of the Nonlinear Stochastic Fokker--Planck Equation

Variational inequalities and smooth-fit principle for singular stochastic control problems in Hilbert spaces

Discrete-Time Approximation of Risk-Averse Control Problems for Diffusion Processes

A description based on optimal transport for a class of stochastic McKean-Vlasov control problems

Stochastic optimal control problems with delays in the state and in the control via viscosity solutions and applications to optimal advertising and optimal investment problems

Approximation of an optimal control problem for the time-fractional Fokker-Planck equation

Optimal control of stochastic delay differential equations and applications to path-dependent financial and economic models

Analysis and Optimal Velocity Control of a Stochastic Convective Cahn–Hilliard Equation

Asymptotic optimality of a class of controlled non-Markov processes