The Method of Statistical Linearization for Non-Linear Stochastic Vibrations
In this paper we discuss certain questions related to the method of statistical linearization as an approximation to the response of non-linear systems. We discuss its second moment properties, its relationship to parameter estimation for linear systems as well as possible connections with recursive estimators. Finally, we show by example that for systems with randomly varying coefficients statistical linearization will yield sample properties that differ significantly from the sample solution properties of the original non-linear system.
Unable to display preview. Download preview PDF.
- 1.R.C. Booton, The Analysis of Non-linear Control Systems with Random Inputs, Proc. MRI Symposium on Nonlinear Circuits, Polytechnic Inst. of Brooklyn, 1953, p. 341–344.Google Scholar
- 4.Y. Sunahara, et al., Statistical Studies in Nonlinear Control Systems, Nippon Printing Co., Osaka, Japan, 1962.Google Scholar
- 8.N.C. Hampl, G.I. Schuëller, Probability Densities of the Response of Non-linear Structures Under Stochastic Dynamic Excitation, Proc. U.S.Austria Seminar on Stoch. Struc. Dynan. Florida Atlantic Univ., Boca Rato, Fl., May 1987.Google Scholar
- 9.T. Kailath, Linear Systems, (Section 9. 1 ), Prentice-Hall, New Jersey, 1980.Google Scholar
- 10.F. Kozin, H.G. Natke, System Identification Techniques, Struct. Safety, Vol. 3, 1986, p. 269–316.Google Scholar
- 13.F. Kozin, Comments Upon the Technique of Statistical Linearization, Proceedings Annual Meeting Japan Assoc. Auto. Contr. Eng., Osaka, 1965 ( Special Lecture).Google Scholar
- 14.A. Bruckner, Y.K. Lin, Generalization of the Equivalent Linearization Method for Nonlinear Random Vibration Problems, Center for Appl. Stoch. Res. Report CAS-86–2, Florida Atlantic Univ., Boca Raton, April, 1986.Google Scholar