Local controllability of a free-boundary problem for a class of one-dimensional degenerate parabolic equations

Lingyang Liu · 2025

Read Local controllability of a free-boundary problem for a class of one-dimensional degenerate parabolic equations by Lingyang Liu (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2503-11929.txt.

Introduction

Lingyang Liu studies whether a moving boundary heat problem can be steered to rest, in a 2025 mathematics paper. The problem is a class of one dimensional equations where the heat spread weakens at a point, and the edge of the region moves freely. Control acts only on part of the region, through a distributed input.

Liu proves a local result: for any chosen end time, if the starting state is small enough, some control drives the state exactly to rest at that time. The proof uses a fixed point theorem of Schauder, joined with careful bounds on solutions of the weakened heat equation and on the control functions. Such results matter for processes where both heat and the boundary itself must be managed.

Catalog record

CREATORLingyang Liu
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE97 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2503-11929

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Local controllability of a free-boundary problem for a class of one-dimensional degenerate parabolic equations Lingyang Liu Abstract This paper is devoted to a study of the controllability of a free-boundary problem for a class of one-dimensional degenerate parabolic equations with distributed controls, locally supported in space. We prove that for any T > 0 T>0 , if the initial state is sufficiently small, there exists a control that drives the state exactly to rest at time t = T t=T . The proof is based on Schauder’s fixed point theorem, combined with appropriate estimates for solutions to degenerate parabolic equations and for control functions. Key Words. Controllability, Carleman estimate, free-boundary problem, 1D degenerate parabolic equation 1 Introduction and main result In real-world applications, many physical phenomena are modeled by degenerate parabolic equations, such as population genetics, fluid dynamics, and vision (see [ 9 ] for a more detailed introduction). Due to these wide-ranging applications, control problems for such equations have received significant attention. Different from nondegenerate parabolic equations, a key feature of degenerate problems is that the diffusion coefficient in the principal operator fails to maintain uniform ellipticity. Consequently, new tools have been developed to study related controllability and observability issues. For instance, novel Carleman-type estimates with adapted weight functions, distinct from those for nondegenerate parabolic equations, have been employed to derive observability inequalities for the corresponding adjoint problems (see, e.g., [ 1 , 7 ] ). To date, control issues concerning different aspects of degenerate parabolic equations in cylindrical domains have been extensively studied. The investigation mainly focuses on the one-dimensional case. Gueye [ 22 ] investigated the boundary null controllability of the one-dimensional degenerate heat equation by the transmutation method, when the control acts through the part of the boundary where degeneracy occurs. Cannarsa et al. [ 6 ] discussed the null controllability of the one-dimensional degenerate heat equation using spectral analysis and the moment method, when the degeneracy occurs inside the domain. In addition, Araújo et al. [ 2 ] recovered the boundary null controllability for the degenerate heat equation by analyzing the asymptotic behaviour of an eligible family of state-control pairs and solving the corresponding singularly perturbed internal null controllability problems. Regarding other aspects, Cannarsa et al. [ 10 ] studied the null controllability cost for a one-dimensional weakly degenerate parabolic equation governed by a boundary control acting at the degeneracy point by means of the moment method. Later, they established a result for strongly degenerate parabolic equations controlled either by a boundary control or by a locally distributed control in [ 11 ] . Besides, Benoit et al. [ 4 ] investigated the null controllability of a class of linear one-dimensional strongly degenerate parabolic equations with measurable coefficients by applying the flatness approach. We also mention that Wang and Du [ 28 ] proved the null controllability of a class of semilinear one-dimensional weakly degenerate parabolic equations involving convection terms. Their approach relies on Carleman estimates for the linearized system and a fixed-point argument. As we know, controllability of degenerate parabolic equations is dependent on the degree of degeneracy. If the degeneracy is too strong, systems may not be null controllable. In such cases, Wang investigated the approximate controllability of a class of linear and semilinear degenerate systems in [ 25 ] and [ 26 ] , respectively. In [ 27 ] , Wang and Du also established the approximate controllability result for a class of semilinear degenerate systems where the convection terms are dominated by the diffusion terms. In practice, many dynamical processes evolve in domains with moving boundaries. A classic example is the changing interface of an ice-water mixture under rising temperatures. Over the last few decades, researchers have studied the stability of degenerate parabolic equations in non-cylindrical domains. For example, M. Bertsch et al. [ 5 ] analyzed the behavior of solutions near the vertex (0,0) for a degenerate parabolic equation in a special class of non-cylindrical domains. In [ 20 ] , Gao et al. investigated the asymptotic behavior of solutions to the degenerate heat equation in non-cylindrical domains with linear moving boundaries. However, to the best of our knowledge, controllability problems for degenerate parabolic equations with moving boundaries (or a free-boundary) have not been addressed. Let T>0 , L_{0}>0 and 0<a<b<L_{\ast}<L_{0} be given. For any function L\in C^{1}([0,T]) with L(t)>0 , set Q_{L}:=\{(x,t):x\in(0,L(t)),t\in(0,T)\} . Write \omega=(a,b) and denote by \chi_{\omega} the characteristic function of \omega . For any 0\leq\alpha<2 , we consider a free-boundary problem for degenerate parabolic systems of the form \left\{\begin{array}[]{ll}y_{t}-(x^{\alpha}y_{x})_{x}=1_{\omega}v,&(x,t)\in Q_{L},\\[5.69054pt] \left\{\begin{array}[]{ll}y(0,t)=0,&\text{if}\quad\alpha\in[0,1),\\[5.69054pt] \lim\limits_{x\rightarrow 0^{+}}x^{\alpha}y_{x}(x,t)=0,&\text{if}\quad\alpha\in[1,2),\\[5.69054pt] \end{array}\right.&\\[5.69054pt] y(L(t),t)=0,&t\in(0,T),\\[5.69054pt] y(x,0)=y_{0}(x),&x\in(0,L_{0}),\\[5.69054pt] L(0)=L_{0},\end{array}\right. (1.1) with the additional boundary condition L^{\prime}(t)=-L^{\alpha}(t)y_{x}(L(t),t),\quad t\in(0,T). (1.2) Here, y is the state variable and v is a control; v acts on the system at any time via the nonempty open set \omega . As observed, ( 1.1 ) degenerates at x=0 for \alpha>0 . More precisely, the degeneracy is weak if 0<\alpha<1 , while strong if \alpha\geq 1 . Note that in ( 1.1 ), a portion of the domain boundary is not predetermined but evolves as part of the solution (referred to as a free boundary), which differs fundamentally from non-cylindrical domain problems where boundary evolution is prescribed. Therefore, to identify the state (L,y) , additional information is needed (termed a free boundary condition) and this is provided by ( 1.2 ). In many areas, this condition is completely natural. For example, consider dissipative systems derived from conservation laws: y_{t}-\text{div}(k(x,t)\nabla y)=0, (1.3) where y denotes the temperature and k\nabla y(x,t) represents the heat flux. If k=1 , obviously, ( 1.3 ) becomes the classical heat equation and a free boundary condition is given by L^{\prime}(t)=-y_{x}(L(t),t),\quad t\in(0,T). Whereas, if the heat conduction coefficient k depends on the degeneracy parameter in a manner as x^{\alpha} , then ( 1.2 ) is proposed. In particular, when \alpha=0 , ( 1.2 ) reduces to the above boundary condition. It is well known that free-boundary problems for parabolic equations serve as mathematical models for a variety of phenomena in nature, such as phase transitions in materials science [ 18 ] , tumor growth in biomedical modeling [ 19 ] , and gas flow through porous media [ 24 ] , among others. In practical applications, ( 1.1 ) may describe the dynamical behavior of a diffusion system where, at one end of the material ( x=0 ), the diffusion coefficient approaches zero, reflecting the cessation of heat or mass transfer at that boundary (e.g., due to perfect insulation or material depletion), while the interface position L(t) evolves over time, governed by energy or mass conservation laws. Let us mention some previous works on the controllability of free-boundary problems for non-degenerate parabolic equations. In [ 17 ] and [ 16 ] , Fernández-Cara et al. investigated the local null controllability of free boundary problems for the linear and semilinear 1D heat equations, respectively. Wang et al. [ 29 ] considered a free boundary problem for a class of quasi-linear 1D parabolic equations and established a local null controllability result. V. Costa et al. [ 13 ] analyzed the local null controllability of a free-boundary problem for the 1D heat equation with local and nonlocal nonlinearities. Concerning the controllability to the trajectories of

Read the complete text