The hierarchical gradient based iterative identification algorithm for multi-input output-error autoregressive systems

Abstract

This paper deals with identification problems for multi-input output error autoregressive (OEAR) systems. The main contribution of this paper is to propose iterative algorithms for multi-input OEAR systems using the hierarchical identification principle. Firstly, a gradient based iterative (GI) identification algorithm is given for multi-input OEAR systems as a comparison. Secondly, using the hierarchical identification principle, we present a hierarchical gradient based iterative (H-GI) identification algorithm. Finally, one example is provided to test the proposed algorithms, the simulation results indicate that the H-GI identification algorithm can produce more accurate parameter estimates than the GI identification algorithm.