Estimating parameters by autosynchronization with dynamics restrictions.

We suggest a general approach to parameter estimation using autosynchronization with some restrictions on system dynamics. This parameter identification method can be extended to estimate parameters from a scalar time series. Furthermore, we propose an average filter method to suppress the influence of noise on parameter estimation. Some limits and extensions of the autosynchronization method are given as well. Several examples are presented to illustrate all methods suggested.