The robust multi‐innovation estimation algorithm for Hammerstein non‐linear systems with non‐Gaussian noise

Abstract

The characteristic of the external noise has significant influences on system modelling and identification, and the assumption that the noise follows the Gaussian distribution may be invalid due to realistic reasons. This paper discusses the identification issue of Hammerstein non-linear systems with non-Gaussian noise and presents a robust gradient algorithm. The algorithm is derived based on the logarithmic cost function of continuous mixed pnorm of prediction errors, which takes into account each p-norm of errors for 1 ⩽ p ⩽ 2. The gain at each recursive step adapts to the data quality so that the algorithm has good robustness to non-Gaussian noise. To improve the estimation accuracy, a robust multiinnovation gradient algorithm is proposed by using the multi-innovation identification theory. Two examples are provided to exhibit the validity of the proposed algorithms. This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.

Modified particle filtering‐based robust estimation for a networked control system corrupted by impulsive noise

“…Although the identification of the systems with non‐Gaussian noise has been reported in the open literature, most of them focused on the case where only the system output was corrupted by outliers 35‐37 . However, in the NCSs, both the output of the dynamic system (the process output) and the output of the communication channel (the communication output) may be corrupted by non‐Gaussian noise because of sensor failures, malicious attacks, and network congestion 38,39 .…”