Parameter Estimation of Bilinear State-Space Systems With Nonlinear Input via Enhanced Nadam Algorithm by Line Search Method

Abstract

This article investigates the problem of parameter estimation for bilinear state-space systems with nonlinear input. An innovative approach that combines the Nesterov-accelerated adaptive moment estimation algorithm with a line search strategy is proposed to address such complex systems. The proposed algorithm uses the backtracking line search method to dynamically select an appropriate step-size, thereby enhancing estimation efficiency. The effectiveness of the proposed algorithm is demonstrated through simulation experiments.