A novel generalized minimum variance (GMV) system identification algorithm
is developed, and its performance is gauged against the established general
ized least squares (GLS) estimation algorithm. The emphasis of the proposed
GMV algorithm is on the rigorous treatment of measurement noise for dynami
cal system identification. A careful analysis of the measurement situation
on hand yields a novel fixed-point calculation-based parameter estimation a
lgorithm. The novel and established algorithms are compared in carefully pe
rformed and reproducible experiments which include measurement noise. Diffe
rences are apparent under small (measurement) sample operation, whereas und
er sufficient excitation, the algorithms produce statistically similar resu
lts.