Interval variable step-size spline adaptive filter for the identification of nonlinear block-oriented system
- 77 Downloads
In order to improve the convergence speed of the nonlinear spline adaptive filter (SAF) in the identification of block-oriented systems, an interval variable step-size algorithm is proposed. Traditional SAF algorithm uses constant step size during iteration, leading to a contradiction between convergence speed and steady-state accuracy. In this paper, a new kind of variable step-size algorithm is proposed, fully considering the particularity of spline interpolation in the nonlinear part of the block-oriented model. The step size of each interpolation interval is independent from that of other intervals, and it is dominated by the correlated squared error which is evaluated by an exponential-weighted averaging (EWA) process. In this paper, the independent step size in each interpolation interval is also updated through an EWA process of the correlated error. The effects of the parameters on the convergence performance of the proposed strategy have been theoretically analyzed and verified by simulations. Finally, some numerical simulations have confirmed that the proposed interval variable step-size approach can significantly improve the convergence speed as well as reduce the steady-state error compared with the traditional SAF and the existing variable step-size SAF algorithms.
KeywordsSpline adaptive filter Block-oriented model Nonlinear identification Variable step size
This research was supported by the National Natural Science Foundation of China (Nos. 51705396, 51835009) and the Postdoctoral Science Foundation of China (No. 2018T111047).
Compliance with ethical standards
Conflict of interest
The authors declare that they have no conflict of interest concerning the publication of this manuscript.
- 1.Billings, S.A.: Identification of nonlinear systems—a survey. In: IEE Proceedings D (Control Theory and Applications), vol. 127, pp. 272–285. IET (1980)Google Scholar
- 15.Liu, W., Principe, J.C., Haykin, S.: Kernel Adaptive Filtering: A Comprehensive Introduction, vol. 57. Wiley, New York (2011)Google Scholar
- 23.Scardapane, S., Scarpiniti, M., Comminiello, D., Uncini, A.: Diffusion spline adaptive filtering. In: 2016 24th European Signal Processing Conference (EUSIPCO), pp. 1498–1502. IEEE (2016)Google Scholar
- 29.Peng, S., Wu, Z., Zhang, X., Chen, B.: Nonlinear spline adaptive filtering under maximum correntropy criterion. In: TENCON 2015-2015 IEEE Region 10 Conference, pp. 1–5. IEEE (2015)Google Scholar
- 33.Patel, V., Comminiello, D., Scarpiniti, M., George, N.V., Uncini, A.: Design of hybrid nonlinear spline adaptive filters for active noise control. In: 2016 International Joint Conference on Neural Networks (IJCNN), pp. 3420–3425. IEEE (2016)Google Scholar
- 37.Zhang, S., Zhang, J., Han, H., Zhang, S., Zhang, J., Han, H.: Robust variable step-size decorrelation normalized least-mean-square algorithm and its application to acoustic echo cancellation. IEEE/ACM Transactions on Audio, Speech and Language Processing (TASLP) 24(12), 2368–2376 (2016)CrossRefGoogle Scholar