A new algorithm is introduced to identify differential equation models for linear and non-linear multiple-input multiple-output systems from frequency response data using a weighted complex orthogonal estimator. The estimation procedure is progressive beginning with the estimation of the linear terms and then sequentially adding higher-order non-linear terms to build up the model. Simulated examples are included to demonstrate the performance of the new algorithm.
UnbehauenH.RaoG. P.Identification of Continuous Systems, System and Control Series, 1987 (North-Holland, Amsterdam).
4.
UnbehauenH.RaoG. P.Continuous time approaches to system identification—a survey. Automatica, 1990, 26 (1), 23–35.
5.
YoungP. C.Parameter estimation of continuous time models—a survey. Automatica, 1981, 17, 23–29.
6.
TsangK. M.BillingsS. A.Reconstruction of linear and nonlinear continuous time models from discrete time sampled data systems. Mech. Systems Signal Processing, 1992, 6 (1), 69–84.
7.
SwainA. K.BillingsS. A.Weighted complex orthogonal estimator for identifying linear and nonlinear continuous time models from generalised frequency response functions. Mech. Systems Signal Processing, 1998, 12 (2), 269–292.
8.
BedrosianE.RiceS. O.The output properties of Volterra systems (nonlinear systems with memory) driven by harmonic and Gaussian inputs. Proc. IEEE, 1971, 59, 1688–1707.
9.
10.
BillingsS. A.VoonW. S. F.Correlation based model validity tests for nonlinear models. Int. J. Control, 1983, 44, 235–244.
11.
BillingsS. A.ZhuQ.A structure detection algorithm for nonlinear dynamic rational model. Int. J. Control, 1994, 59, 1439–1463.
12.
LinkensD. A.Analytical solution of large number of mutually coupled nearly sinusoidal oscillators. IEEE Trans. Circuits Systems, 1974, CAS-21 (2), 294–300.
