Enhanced FDLS algorithm for parameter-based system identification for automatic testing

Theoretical analyses of (1) the frequency domain least square (FDLS) algorithm using an impulse as the input, (2) the FDLS algorithm using white Gaussian noise as the input, and (3) the FDLS algorithm using sweep frequency inputs of pure cosines and sines each have advantages depending upon the environment in which they are applied. In comparing these algorithms in the specific application of characterizing system components and automatically testing those components, the authors have discovered that the FDLS algorithm using sweep frequencies of pure cosine and sine inputs offers significant advantages over the other two approaches. Furthermore, the authors propose a modification to this algorithm that significantly enhances its ability to identify system poles in this specific environment