Improved Variance Estimates of FRF Measurements in the Presence of Nonlinear Distortions Via Overlap

The frequency response function (FRF) is a common nonparametric modeling tool in many practical engineering problems used for obtaining insight in the device under test. However, the device often behaves nonlinearly. When nonlinearities are detected, the user wants to find out how large these are with respect to the measurement noise. In this paper, we describe a method, based on an overlap technique and periodic excitations, that accurately estimates the FRF, the level of nonlinear distortion, and the measurement noise using only two periods and two random phase realizations of the input signal.