1: \begin{abstract}

2:   The synchronization of dynamical systems is a method that allows two

3:   systems to have identical state trajectories, appart from an error

4:   converging to zero. This method consists in an appropriate unidirectional coupling

5:   from one system (drive) to the other (response).

6:   % Even though the state trajectory of a chaotic dynamical system may

7:   % change drastically given an arbitrarily small change in the initial

8:   % condition, two identical chaotic systems can be synchronized with an

9:   % unidirectional coupling between them. 

10:   This requires that the response system shares the same dynamical

11:   model with the drive.  For the cases where the drive is unknown,

12:   Chen proposed in 2002 a method to adapt the response system such

13:   that synchronization is achieved, provided that (1) the response

14:   dynamical model is linear with a vector of parameters, and (2) there

15:   is a parameter vector that makes both system dynamics identical.

16:   %

17:   However, this method

18:   has two limitations: first, it does not scale well for complex

19:   parametric models (e.g., if the number of parameters is greater than

20:   the state dimension), and second, the model parameters are not

21:   guaranteed to converge, namely as the synchronization error

22:   approaches zero. This paper presents an adaptation law 

23:   addressing these two

24:   limitations. Stability and convergence proofs, using Lyapunov's

25:   second method, support the proposed adaptation law. Finally,

26:   numerical simulations illustrate the advantages of the proposed method, namely

27:   showing cases where the Chen's method fail, while the proposed one does not.

28: \end{abstract}

29: