Optimal Design of Hammerstein Cubic Spline Filter for Nonlinear System Modeling Based on Snake Optimizer

Abstract

This work develops a new class of Hammerstein adaptive filters that contains a memory-less non-linear system followed by a linear adaptive filter, where the non-linear system comprises an adjustable look-up table (LUT) and a spline interpolator. This paper's first effort has been to employ the meta-heuristic algorithm (MHA) to the Hammerstein spline adaptive filter (HSAF), where it concurrently updates the weights of spline control points and linear filter based on the estimation problem. A novel MHA called snake optimiser algorithm (SOA) is used to enhance the assurance of convergence, estimated parameter accuracy and steady-state results. The presented experimental results indorse that the proposed SOA-based HSAF (SOA-HSAF) design exhibits more robust performance in dealing with higher degree non-linear systems under the Gaussian and Non-Gaussian circumstances compared to contemporary design methods like classical, some standard MHAs and other researchers reported techniques. The achieved simulation results are validated using a TMS320C6713 digital signal processor.