A randomized integral error criterion for parametric identification of dynamic models of mechanical systems

This paper proposes a new approach to the identification of reduced order m odels for complex mechanical vibration systems. Parametric identification i s commonly conducted by the regression of time-series data, but when this i ncludes significant unmodelled modes, the error process has a high variance and autocorrelation. In such cases, optimization using least-squares metho ds can lead to excessive parameter bias. The proposed method takes advantag e of the inherent boundedness of mechanical vibrations to design a new regr ession set with dramatically reduced error variance. The principle is first demonstrated using a simple two-mass simulation mode l, and from this a practicable approach is derived. Extensive investigation of the new randomized integral error criterion method is then conducted us ing the example of identification of a quarter-car suspension system. Simul ation results are contrasted with those from comparable direct least-square s identifications. Several forms of the identification equations and error sources are used, and in all cases the new method has clear advantages, bot h in accuracy and consistency of the resulting identification model.