Simple approaches for the identification of Wiener-type nonlinear processes

We propose two simple approaches to identify Wiener-type nonlinear processes. One approach uses two step signals with different magnitudes to activate the process and the other approach uses a rectangular pulse signal and a step signal. These specialized signals can make the identification problem of the nonlinear static function be independent on that of the linear dynamic subsystem, resulting in significantly simplified identification procedure. The unmeasurable intermediate variable and the nonlinear static function can be estimated without any iterative procedures and also, the model structure can be determined in a much simpler manner. Once we obtain the nonlinear static function, we can design a nonlinear controller that linearizes the nonlinear dynamics using the inverse of the nonlinear static function and it can be tuned by usual tuning rules developed for linear processes.