作者: Er-Wei Bai , Jr. Reyland , John
DOI: 10.1016/J.AUTOMATICA.2008.11.020
关键词:
摘要: In this paper, we investigate what constitutes the least amount of a priori information on nonlinearity so that linear part is identifiable in non-Gaussian input case. Under white noise input, three types are considered: quadrant information, point and monotonic information. all cases, identifiability has been established, corresponding nonparametric identification algorithms developed along with their convergence proofs.