Online identification method for Wiener nonlinear systems
摘要
Wiener system is one of the most studied nonlinear systems. Presently, an identification method for Wiener nonlinear systems, using the extremum-seeking optimization technique, is developed. Most of the previous methods on system identification can be established offline. This new identification solution is an online technique. In this identification problem, it is well known that the unknown parameters are not linear with respect to data acquisition. To estimate the phase of linear block, a noise signal of small amplitude is added to the phase estimate. Taking advantage of the non-uniqueness of the solution and using the phase estimate, it is shown that the nonlinear block parameters become linear with respect to the input and output data. Then, the estimates of nonlinearity parameters can be obtained suing least square algorithm. By comparing the output of true nonlinear system and that of the estimated model, the identification of system parameters can be obtained based on online collected measurements. This method is easy to be implemented, and it is characterized by rapid convergence. Several simulation results highlight the performances of the method.