Abstract
In this paper, we consider the problem for optimal control of the sixth-order convective Cahn-Hilliard type equation. The optimal control under boundary condition is given, the existence of an optimal solution to the equation is proved and the optimality system is established.
Similar content being viewed by others
1 Introduction
In past decades, the optimal control of a distributed parameter system has received much more attention in academic field. A wide spectrum of problems in applications can be solved by the methods of optimal control such as chemical engineering and vehicle dynamics. Many papers have already been published to study the control problems for nonlinear parabolic equations, for example, [1]–[9] and so on.
The Cahn-Hilliard (CH) equation is a type of higher order nonlinear parabolic equation, it models many interesting phenomena in mathematical biology, fluid mechanics, phase transition, etc. The fourth-order convective Cahn-Hilliard (FCCH) equation arises naturally as a continuous model for the formation of facets and corners in crystal growth. Many papers have been devoted to CH equation and FCCH equation, see, for example, [10]–[15]. In [16], Savina et al. derived a sixth-order convective Cahn-Hilliard (SCCH) equation
for the description of a growing crystalline surface with small slopes that undergoes faceting. Here, is the slope of a surface and δ is proportional to the deposition strength of an atomic flux. Recently, by an extension of the method of matched asymptotic expansions that retains exponentially small terms, Korzec et al.[17] derived a new type of stationary solutions of the one-dimensional sixth-order Cahn-Hilliard equation. In [18], the existence and uniqueness of weak solutions to equation (1) with periodic boundary conditions were established in . Furthermore, a numerical study showed that the solution behave similarly to that for the better known convective Cahn-Hilliard equation. We also noticed that some investigations of SCCH equation were studied, such as in [13], [19].
In this article, suppose that κ is a positive constant, S is a real Hilbert space of observations, is an operator, which is called the observer. We are concerned with the distributed optimal control problem
subject to
on an interval for , where is the cost function associated with the control system. The control target is to match the given desired state in -sense by adjusting the body force w in a control volume in the -sense. On the other hand, we assume that the initial function of (3) has zero mean, i.e., , then it follows that for .
Now, we introduce some notations that will be used throughout the paper. For fixed , let be an open set with positive measure, and , let and be dual spaces of V and H. Then we get
Each embedding is dense. The extension operator which is called the controller is introduced as
We supply H with the inner product and the norm , and we define a space as
which is a Hilbert space endowed with common inner product.
This paper is organized as follows. In the next section, we give some preparations and establish the existence and uniqueness of a global solution for problem (3). In Section 3, we consider the optimal control problem and prove the existence of an optimal solution. In Section 4, the optimality conditions are showed and the optimality system is derived.
In the following, the letters c, () will always denote positive constants different in various occurrences.
2 Global existence and uniqueness of weak solution
In this section, we prove the existence and uniqueness of a weak solution for problem (3).
Definition 2.1
For all , a function is called a weak solution to problem (3), if
Now, we give Lemma 2.1, which ensures the existence of a unique weak solution to problem (3).
Lemma 2.1
Suppose, , . Then problem (3) admits a unique weak solution.
Proof
The Galerkin method is applied to the proof. Denote as a differential operator, let denote the eigenfunctions of the operator . For , define the discrete ansatz space by
Let require in H hold true.
By analyzing the limiting behavior of sequences of a smooth function , we can prove the existence of a weak solution to problem (3).
Performing the Galerkin procedure for (3), we obtain
Obviously, the equation of (5) is an ordinary differential equation, and according to ODE theory, there exists a unique solution to problem (5) in the interval . What we should do is to show that the solution is uniformly bounded when . We need also to show that the times there are not decaying to 0 as .
Then, we shall prove the existence of a solution for problem (5). Setting
Differentiating with respect to time and integrating by parts, we obtain
Using Poincaré’s inequality, we have
Hence
Since is the control item, we can assume , where M is a positive constant. Then, by Gronwall’s inequality, we get
A simple calculation shows that
Therefore
By the Sobolev embedding theorem, we get
Multiplying the equation of (5) by , integrating with respect to x on , we obtain
A simple calculation shows that
Hence, we have
that is,
Therefore
Multiplying the equation of (5) by , integrating with respect to x on , we deduce that
Note that
By Nirenberg’s inequality and (7), we have
and
Then
We also have
and
Summing up, we have
Therefore
Hence, we have
In addition, we prove a uniform bound on a sequence . Noticing that
By the Sobolev embedding theorem, we have . Therefore
Then, we immediately conclude
Based on the above discussion, we obtain . It is easy to check that is continuously embedded into which denotes the space of continuous functions. We conclude the convergence of a subsequence, denoted by , weak into , weak-star in (by [20], Lemma 4]) and strong in to functions . Since the proof of uniqueness is similar as the proof of Theorem 2 of [17], we omit it.
Then, we complete the proof. □
Now, we will discuss the relation among the norm of a weak solution and the initial value and the control item.
Lemma 2.2
Suppose, , . Then there exist positive constantsandsuch that
Proof
Setting
passing to the limit in (6), we obtain
Using Gronwall’s inequality, we get
A simple calculation shows that
Passing to the limit in (11), we obtain
Therefore
On the other hand, we have
Note that
By the Sobolev embedding theorem, we have . Therefore
Then, we immediately conclude
By (19), (20), (21), (22) and the definition of extension operator B, we obtain (18). Hence, Lemma 2.2 is proved. □
3 Optimal control problem
In this section, we consider the optimal control problem associated with the sixth-order convective Cahn-Hilliard equation and prove the existence of an optimal solution.
In the following, we suppose that is a Hilbert space of control variables, we also suppose that is the controller and a control , . Consider (3), by virtue of Lemma 2.1, we can define the solution map of into . The solution is called the state of control system (3). The observation of the state is assumed to be given by Cu. The cost function associated with control system (3) is given by
The optimal control problem about (3) is
where satisfies (3).
Let and . We define an operator , where
Here is an operator from V to . Hence, we write (24) in the following form:
Now, we give Theorem 3.1 on the existence of an optimal solution to the sixth-order convective Cahn-Hilliard equation.
Theorem 3.1
Suppose. Then there exists an optimal control solutionto problem (3).
Proof
Suppose that satisfies the equation . In view of (23), we deduce that
By Lemma 2.1, we obtain
Therefore,
As the norm is weakly lower semi-continuous, we achieve that J is weakly lower semi-continuous. Since for all , , there exists defined by
which means the existence of a minimizing sequence in X such that
From (25), there exists an element such that when ,
Then, using (26), we get
By the definition of and the compactness of embedding and , we find from (26) and the results of Lemma 2.1 that strongly in and strongly in when .
Since the sequence converges weakly and is bounded in , based on the embedding theorem, we can obtain that is also bounded.
Because strongly in as , by [20], Lemma 4], we know that is bounded. Because strongly in when , we know that is bounded too.
Using (27) again, we derive that
By (26), we deduce that
For , a simple calculation shows that
For , we get
Then, we immediately obtain
We also have the following inequality:
In view of the above discussion, we get
As is known, , we derive that . Since weakly in , we get weakly when . Thus, we obtain
which means . Therefore, we obtain
So, there exists an optimal solution to problem (3). Then, we complete the proof of Theorem 3.1. □
4 Optimality conditions
It is well known that the optimality conditions for u are given by the variational inequality
where denotes the Gâteaux derivative of at .
The following lemma is essential in deriving necessary optimality conditions.
Lemma 4.1
The mapofintois weakly Gâteaux differentiable at, and such the Gâteaux derivative ofatin the direction, say, is a unique weak solution of the following problem:
Proof
Let , and u be the solutions of (3) corresponding to and w, respectively. Then we prove the lemma in the following two steps.
Step 1, we prove strongly in as . Let , then
Taking the scalar product of (31) with q, a simple calculation shows that
By Lemmas 2.1-2.2 and the Sobolev embedding theorem, we get
In addition, a simple calculation shows that
Hence
that is,
Taking the scalar product of (31) with , a simple calculation shows that
that is,
Adding (32)-(33) together gives
Using Gronwall’s inequality, it is easy to see that as . Therefore, strongly in as .
Step 2, we prove that strongly in . Now, we rewrite (31) in the following form:
We can easily verify that the above problem possesses a unique weak solution in . On the other hand, it is easy to check that the linear problem (30) possesses a unique weak solution . Let , thus r satisfies
Taking the scalar product of the equation of the above problem with r, we get
Noticing that
where . Taking the scalar product of (34) with and , respectively, a simple calculation shows that
Using Gronwall’s inequality, we obtain
On the other hand, we have and strongly in as , then
Noticing that strongly in as , thus
Therefore
Using the same method as above, we get
Summing up, we obtain
Using Gronwall’s inequality, it is easy to check that is strongly convergent to z in . Then, Lemma 4.1 is proved. □
As in [1], [4], we denote Λ= canonical isomorphism of S onto , where is the dual space of S. By calculating the Gateaux derivative of (25) via Lemma 4.1, we see that the cost is weakly Gateaux differentiable at u in the direction . Then, , (29) can be rewritten as
where z is the solution of (30).
Now we study the necessary conditions of optimality. To avoid the complexity of observation states, we consider the two types of distributive and terminal value observations.
-
1.
Case of .
In this case, and (35) may be written as
We introduce the adjoint state by
According to Lemma 2.1, the above problem admits a unique solution (after changing t into ).
Multiplying both sides of (37) (with ) by z, using Lemma 4.1, we get
and
Then, we obtain
Hence, (36) may be written as
Therefore we have proved the following theorem.
Theorem 4.1
We assume that all the conditions of Theorem 3.1hold. Let us suppose that. The optimal control w is characterized by the system of two PDEs and an inequality: (3), (37) and (38).
-
2.
Case of .
In this case, we observe , . The associated cost function is expressed as
Then, , the optimal control w for (39) is characterized by
We introduce the adjoint state by
According to Lemma 2.1, the above problem admits a unique solution (after changing t into ).
Let us set in the above equations and scalar multiply both sides of the first equation of (41) by and integrate from 0 to T. A simple calculation shows that (40) is equivalent to
Then, we have the following theorem.
Theorem 4.2
We assume that all the conditions of Theorem 3.1hold. Let us suppose that. The optimal control w is characterized by the system of two PDEs and an inequality: (3), (41) and (42).
References
Ryu S-U, Yagi A: Optimal control of Keller-Segel equations. J. Math. Anal. Appl. 2001, 256: 45-66. 10.1006/jmaa.2000.7254
Tian L, Shen C: Optimal control of the viscous Degasperis-Procesi equation. J. Math. Phys. 2007., 48(11): 10.1063/1.2804755
Zhao X, Liu C: Optimal control problem for viscous Cahn-Hilliard equation. Nonlinear Anal. 2011, 74: 6348-6357. 10.1016/j.na.2011.06.015
Lions JL: Optimal Control of Systems Governed by Partial Differential Equations. Springer, Berlin; 1971.
Yong J, Li X: Optimal Control Theory for Infinite Dimensional Systems. Birkhäuser, Basel; 1995.
Zhao X, Liu C: Optimal control for the convective Cahn-Hilliard equation in 2D case. Appl. Math. Optim. 2014, 70: 61-82. 10.1007/s00245-013-9234-0
Zheng J: Time optimal controls of the Lengyel-Epstein model with internal control. Appl. Math. Optim. 2014.
Zheng J, Liu J, Liu H: State-constrained optimal control of phase-field equations with obstacle. Bound. Value Probl. 2013., 2013: 10.1186/1687-2770-2013-234
Zheng J, Wang Y: Time optimal controls of the Fitzhugh-Nagumo equation with internal control. J. Dyn. Control Syst. 2013, 4(19):483-501. 10.1007/s10883-013-9190-7
Goldstein GR, Miranville A, Schimperna G: A Cahn-Hilliard model in a domain with non-permeable walls. Physica D 2011, 240: 754-766. 10.1016/j.physd.2010.12.007
Dlotko T, Kania MB, Sun C:Analysis of the viscous Cahn-Hilliard equation in . J. Differ. Equ. 2012, 252: 2771-2791. 10.1016/j.jde.2011.08.052
Miranville A: Long-time behavior of some models of Cahn-Hilliard equations in deformable continua. Nonlinear Anal., Real World Appl. 2001, 2: 273-304. 10.1016/S0362-546X(00)00104-8
Golovin AA, Nepomnyashchy AA, Davis SH, Zaks MA: Convective Cahn-Hilliard models: from coarsening to roughening. Phys. Rev. Lett. 2001, 86: 1550-1553. 10.1103/PhysRevLett.86.1550
Watson SJ, Otto F, Rubinstein BY, Davis SH: Coarsening dynamics of the convective Cahn-Hilliard equation. Physica D 2003, 178: 127-148. 10.1016/S0167-2789(03)00048-4
Zhao X, Liu C: Optimal control of the convective Cahn-Hilliard equation. Appl. Anal. 2013, 92: 1028-1045. 10.1080/00036811.2011.643786
Savina TV, Golovin AA, Davis SH, Nepomnyashchy AA, Voorhees PW: Faceting of a growing crystal surface by surface diffusion. Phys. Rev. E 2003., 67: 10.1103/PhysRevE.67.021606
Korzec MD, Evans PL, Munch A, Wagner B: Stationary solutions of driven fourth and sixth-order Cahn-Hilliard-type equations. SIAM J. Appl. Math. 2008, 69(2):348-374. 10.1137/070710949
Korzec MD, Rybka P: On a higher order convective Cahn-Hilliard type equation. SIAM J. Appl. Math. 2012, 72: 1343-1360. 10.1137/110834123
Korzec MD, Nayar P, Rybka P: Global weak solutions to a sixth order Cahn-Hilliard type equation. SIAM J. Math. Anal. 2012, 44: 3369-3387. 10.1137/100817590
Simon J: Nonhomogeneous viscous incompressible fluids: existence of velocity, density, and pressure. SIAM J. Math. Anal. 1990, 21(5):1093-1117. 10.1137/0521061
Author information
Authors and Affiliations
Corresponding author
Additional information
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
All authors contributed equally to the writing of this paper. All authors read and approved the final manuscript.
Rights and permissions
Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (https://creativecommons.org/licenses/by/4.0), which permits use, duplication, adaptation, distribution, and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.
About this article
Cite this article
Zhao, X., Duan, N. Optimal control of the sixth-order convective Cahn-Hilliard equation. Bound Value Probl 2014, 206 (2014). https://doi.org/10.1186/s13661-014-0206-3
Received:
Accepted:
Published:
DOI: https://doi.org/10.1186/s13661-014-0206-3