Separable synchronous auxiliary model adaptive momentum estimation strategy for a time-varying system with colored noise from on-line measurements

Abstract

The primary focus of this article is to explore parameter estimation for time-varying systems affected by colored noise. Based on the attributes of the time-varying system with colored noise under investigation, the original system is separated and two different subsystems are reconstructed. To address the influence of the hidden variables in the system and the time-varying noise signal, we introduce auxiliary models into the reconstructed systems to achieve the separation and synchronization estimation of the time-varying parameters within the system. With the aim of achieving high-precision performance in estimating the time-varying parameters, a separable synchronous auxiliary model adaptive momentum algorithm is presented by introducing bias correction to the momentum and gradient square term and online parameter estimation is implemented. The proposed algorithm is used for estimating the time-varying output error moving average model to verify the performance. The results of simulation experiments illustrate the efficacy of the proposed method for estimating the time-varying system with colored noise. Additionally, the proposed algorithm is extended to a kind of direct current (DC) motor modeling parameter identification problem and shows good tracking performance.