Using the best linear approximation with varying excitation signals for nonlinear system characterization

Since nonlinear (NL) identification is often very time consuming, it is highly desirable to carefully select a well suited model structure prior to the actual identification. In this study the focus is on the detection of the presence of NL feedback. Therefore, the frequency response function (FRF) of the best linear approximation (BLA) is measured at different excitation signal settings. Two properties of the excitation are varied to analyze the nonlinear behavior:
1. Amplitude (RMS value) level
2. DC level (offset or set points)
Besides the measurement of the BLA, we make, at the same time, a non-parametric analysis of the nonlinear distortions. This can be done with either the robust method, or with the fast method. The goal of the study is to find out what is the best strategy to vary the excitation signal.