Abstract

This paper studies the adaptive control problem of the Korteweg-de Vries-Burgers equation. Using the Lyapunov function method, we prove that the closed-loop system including the parameter estimator as a dynamic component is globally stable. Furthermore, we show that the state of the system is regulated to zero by developing an alternative to Barbalat's lemma which cannot be used in the present situation. The closed-loop system is shown to be well posed.

1. Introduction

In this paper, we are concerned with the problem of boundary adaptive control of the KdVB equation: where the viscosity parameter . The dispersion parameter is unknown, and are control inputs, and is an initial state in an appropriate function space. When , the KdVB equation becomes the KdV equation; when , it becomes the Burgers equation.

The problem of control of the Burgers, KdV, and KdVB equations has received extensive attention for several decades [17]. In [2], Liu and Krstic obtained the adaptive control of the Burgers equation. Up to now, it seems not to have many discussions on the adaptive control of the KdV and KdVB equation. In this paper we establish a Barbalat-like lemma [8] and use the Lyapunov function method to prove that the system of the KdVB equation is globally stable under the boundary conditions. Using Banach fixed point theorem, we proved the well-posedness of the KdVB equation under the given boundary condition.

The rest of the paper is organized as follows. We present our main results in Section 2. In Section 3, we establish the alternative to Barbalat’s lemma. In Section 4, we prove that the KdVB equation with the previous adaptive boundary feedbacks is globally stable. By the alternative to Barbalat’s lemma, we show the regulation of the solution. In Section 5, we establish the global existence and uniqueness of the solution with help of the Banach fixed point theorem.

We now introduce some notations used throughout the paper. denotes the usual Sobolev space [9] for any . For , denotes the completion of in , where denotes the space of all infinitely differentiable functions on with compact support in . The norm is defined in the usual way, . The norm on is denoted by . It is easy to see that Let be the Banach space and . We denote by the space of times continuously differentiable functions defined on with values in . We denote by the scalar product of .

2. Main Result

For notational convenience, in what follows, we denote where will be used as estimates of .

Consider the system satisfies the following theorem.

Theorem 2.1. Suppose that , the initial condition , and , . If the problem (2.2) has a global solution , then one has the equilibrium is globally -stable, that is: and is regulated to zero in sense:

3. The Alternative to Barbalat’s Lemma

Recently, the Barbalat’s lemma has more and more important applications in control theory, especially in the adaptive control theory. It is easy to connect with Lyapunov method to analyze the stability and convergence of the system. In this section, we establish the following alternative to Barbalat’s lemma [8].

Lemma 3.1. Suppose that the function defined on satisfies the following conditions:(i) for all ,(ii) is differentiable on and there exists a constant such that ,(iii).
Then one has

Proof. Since, we have is uniformly continuous, and therefore is uniformly continuous.
Setting we have , such that is also uniformly continuous.
By the standard Barbalat’s lemma, we have , then

4. Proof of Stabilization

In this section, we prove our main result by the Lyapunov method. Now we present the proof of Theorem 2.1; first we prove the stability of the system, and then we prove the exisetence and uniqueness of the solution.

Step 1. Stability (2.3). We follow the Lyapunov approach: to this end, we introduce the energy function and the Lyapunov function where is a positive constant. Using (1.1) and integrating by parts, we obtain This leads us to select the adaptive feedback control: where is any positive constant. By this control, we obtain Thus This shows that (2.3) holds.
And we have:

Step 2. Regulation (2.4). To prove (2.4), it suffices to verify conditions (ii) and (iii) of Lemma 3.1.
By (1.2) and (4.7), we obtain Here and in the sequel,  denotes a generic positive constant depending on , which may vary from line to line. Thus condition (iii) of Lemma 3.1 is fulfilled. On the other hand, using Young inequality and noting that , we have which, combining with (4.6), implies condition (ii) of Lemma 3.1.
Hence

5. Well-Posedness

In this section, we use the Banach fixed point theorem to prove that the problem (2.2) is well posed. To this end, for any constant , we first consider the following linear boundary value problem for any fixed We introduce some notation as follows.

If and are two solutions of the problem (5.1) corresponding to and respectively, we set For a general function , we set

Lemma 5.1. If and the initial data , then the problem (5.1) has a unique weak solution satisfying . Moreover, one has where is a positive continuous function.

Proof. We use the standard Galerkin method. This method relies on a number of prior estimates which can be usually obtained by using Gronwall’s inequality.
Step  1. Transformation to a Homogeneous Problem. In order to use the Galerkin method, we transform the problem (5.1) into a homogenous boundary value problem. We first assume that is infinitely differentiable with respect to both and . Set Then it is clear that satisfies the following equation: where
Step  2. Approximate Problem. Set Let be an orthonormal basis in such that each function is in . Since , can be expanded as Set Then we have Consider the following approximate problem: Multiplying (5.12) by and integrating from to , we obtain where Since the matrix is nonsingular, by the classical theory of ordinary differential equations, the linear problem (5.14) has a unique continuously differential solution on . Therefore, the approximate problems (5.12) and (5.13) have a unique solution with .
Step  3. A Priori Estimate on . In what follows, we denote by a positive generic constant, independent of , which may vary from line to line. Multiplying (5.12) by and integrating from to , we have By Gronwall-Bellman’s inequality and using inequality ,
Step  4. A Priori Estimate on . Multiplying (5.12) by and integrating from to , we have which implies
Step  5. Existence and Uniqueness. By (5.17) and (5.19), we deduce that and are bounded in . Consequently, there exists a subsequence of denoted by such that converges to a function in the weak-star topology of .
It is easy to see that is the weak solution of the problem (5.6) satisfying Therefore, for any differentiable function , the problem (5.1) has a unique weak solution: The estimate (5.4) can be proved in the same way as in the proof of (5.19). Finally, the continuous differentiability assumption on can be relaxed by using (5.4). This completes the proof.

Theorem 5.2. For the initial data , the problem (2.2) has a unique solution satisfying .

Proof. Let be any constant and . By Lemma 5.1, the problem (5.1) has a unique solution with . Hence we define the nonlinear mapping by
Set . By (5.4), we deduce that if is small enough, then maps into and is a contractive mapping, where Therefore, by the Banach fixed point theorem, has a unique fixed point . So the problem (2.2) has a unique solution for small enough. Since is also the solution of the linear problem (5.1), by Lemma 5.1, we deduce that . By Theorem 2.1, the solution can be continued to the whole real line, that is, . This completes the proof.

6. Conclusion

We have shown the adaptive boundary stabilization of the KdV-Burgers’ equation by nonlinear boundary control. It seems yet not to be discussed. As for the adaptive case, we establish an extension to Barbalat’s lemma to show the regulation of the solution of the KdV-Burgers’ equation. Also, we prove that the solution of the KdV-Burgers equation is well posed. However, we want to get the stability which implies boundedness of the solution. We will solve this problem in the future.

Acknowledgments

This work is supported by the National Nature Science Foundation of China (no. 11171135, 71073072) and the Nature Science Foundation of Jiangsu (no. BK2010329) and the Natural Science Foundation of the Jiangsu Higher Education Institutions of China (no. 09KJB110003).