Design and Implementation of a Noise Tolerant Polynomial Nonlinear ARX Model using the Averaging Wavelet Method
Khadem Olama Ehsan, Jazayeri-Rad Hooshang · International Journal of Computer Applications · 2011
In this paper, a new nonlinear wavelet identification structure is proposed for high noise resistive soft sensors. This method uses proposedPolynomial Nonlinear Auto Regressive Exogenous Model, which can be solved with linear Gaussian Least Square Method, alongside the Averaging Wavelet Method (AWM) filter. AWM uses the approximation spaces for analyzing the signals and reduce the noise by a mean filtering over subresolutions. Conventional wavelet modeling methods use the detail spaces of the decomposed signal for signal modeling. Theapplication results show that this method can be more accuratein high level noisy environments than the conventional wavelet modeling methods cab tolerate. General Terms Wavelets, Nonlinear Auto Regressive Exogenous Modeling