Wavelet basis expansion-based spatio-temporal Volterra kernels identification for nonlinear distributed parameter systems
C. M. Cheng,X. J. Dong,Z. K. Peng,W. M. Zhang,G. Meng
DOI: https://doi.org/10.1007/s11071-014-1506-y
IF: 5.741
2014-01-01
Nonlinear Dynamics
Abstract:In practical industries, there are many systems belong to nonlinear distributed parameter systems (DPS); unfortunately, modeling of nonlinear DPS is a challenging task because of the infinite-dimensional and nonlinear properties. To model the nonlinear DPS, a spatio-temporal Volterra model is presented with a series of spatio-temporal kernels. It can be considered as a spatial extension of the traditional Volterra model. One question involved in modeling a spatio-temporal functional relationship between the input and output of nonlinear distributed parameter systems using spatio-temporal Volterra series is to identify the spatio-temporal Volterra kernel functions. In addition, in order to derive a low-order model, the Karhunen–Loève (KL) decomposition is used for the time/space separation. The basic routine of the approach is that, first, from the system outputs, KL decomposition is used for the time/space separation, where the spatio-temporal output is decomposed into a few dominant spatial basis functions with temporal coefficients. Second, according to temporal coefficients of outputs under multilevel excitations, the Volterra series outputs of different orders are estimated with the wavelet balance method. Third, the Volterra kernel functions of different orders are separately estimated through their corresponding Volterra series outputs by expanding them with four-order B-spline wavelet on the interval (BSWI). Finally, the spatio-temporal Volterra model can be reconstructed using the time/space synthesis. The simulation studies verify the effectiveness of the presented identification method.