[Paper Review] Minimal time control of exact synchronization for parabolic systems

Lijuan Wang, Qishu Yan|arXiv (Cornell University)|Mar 1, 2018

Stability and Controllability of Differential Equations16 references3 citations

TL;DR

This paper establishes a necessary and sufficient condition for minimal time control of exact synchronization in linear parabolic systems, proving that optimal control exists if and only if the control norm exceeds a critical threshold dependent on the initial state. The key contribution is a duality-based characterization linking minimal time control to minimal norm control problems via a time-optimal control framework for parabolic PDEs with state constraints.

ABSTRACT

This paper studies a kind of minimal time control problems related to the exact synchronization for a controlled linear system of parabolic equations. Each problem depends on two parameters: the bound of controls and the initial state. The purpose of such a problem is to find a control (from a constraint set) synchronizing components of the corresponding solution vector for the controlled system in the shortest time. In this paper, we build up a necessary and sufficient condition for the optimal time and the optimal control; we also obtain how the existence of optimal controls depends on the above mentioned two parameters.

Motivation & Objective

• To address the minimal time control problem for exact synchronization in linear parabolic systems with bounded controls.
• To determine the existence and characterization of optimal controls that synchronize the system components in the shortest possible time.
• To establish a duality between minimal time control and minimal norm control problems for parabolic systems.
• To derive necessary and sufficient conditions for the existence of optimal controls based on the initial state and control bound.
• To analyze how the optimal time and control depend on the initial state and the control norm constraint.

Proposed method

• Formulates a minimal time control problem for a linear parabolic system with Dirichlet boundary conditions and distributed control in a subdomain.
• Defines the control constraint set $\mathcal{U}_M$ with $L^2$-norm bounded by $M > 0$, and the target set $S$ of synchronized states where all components are equal.
• Introduces the operator $D$ to represent the difference between components, so $D\bm{y} = \bm{0}$ characterizes exact synchronization.
• Applies duality theory to relate the minimal time control problem to a minimal norm control problem for the transformed system $\bm{z}_t - \Delta \bm{z} + A\bm{z} = \chi_\omega B\bm{u}$ with initial data $D\bm{y}_0$.
• Uses Pontryagin's maximum principle and time-reversal techniques to derive optimality conditions for the minimal time problem.
• Establishes equivalence between optimal controls of the minimal time problem and those of a minimal norm problem via the function $N(T, \bm{y}_0)$, which denotes the minimal $L^2$-norm of control needed to achieve synchronization by time $T$.

Experimental results

Research questions

• RQ1 Under what conditions does a minimal time control exist for exact synchronization of a parabolic system with a given control bound?
• RQ2 How does the optimal time depend on the initial state and the control norm constraint?
• RQ3 What is the relationship between minimal time control and minimal norm control in the context of parabolic systems?
• RQ4 Can the optimal control for the minimal time problem be characterized via duality with a minimal norm control problem?
• RQ5 What is the precise threshold of the control norm $M$ that guarantees the existence of an optimal control for synchronization?

Key findings

• The optimal time $T(M, \bm{y}_0)$ is finite if and only if $M > M(\bm{y}_0)$, where $M(\bm{y}_0)$ is the minimal $L^2$-norm required to achieve synchronization.
• An optimal control exists for the minimal time problem if and only if the control norm $M$ exceeds the threshold $M(\bm{y}_0)$, which depends on the initial state $\bm{y}_0$.
• The optimal control $\bm{u}^*$ for $(TP)_M^{\bm{y}_0}$ satisfies $\|\bm{u}^*\|_{L^2(0,\infty;L^2(\Omega)^m)} = N(T(M,\bm{y}_0), \bm{y}_0)$, linking the minimal time to a minimal norm problem.
• The optimal time $T(M, \bm{y}_0)$ satisfies $T(M, \bm{y}_0) = T(N(T(M, \bm{y}_0), \bm{y}_0), \bm{y}_0)$, showing a self-consistent duality between time and norm minimization.
• The optimal control $\bm{u}^*$ is nonzero on a positive measure set in $(0, T(M, \bm{y}_0))$, confirming that the control is active throughout the synchronization process.
• The problem is well-posed: if $M > M(\bm{y}_0)$, then $T(M, \bm{y}_0) > 0$ and the optimal control is unique and characterized by the minimal norm control of the transformed system.