Title: Highly efficient online stochastic gradient and sliding window stochastic gradient signal modelling methods for multi-frequency signals

Authors: Guanglei Song; Ling Xu

Addresses: School of Mechanical Technology, Wuxi Institute of Technology, 214122, Wuxi, Jiangsu, China ' School of Internet of Things Technology, Wuxi Vocational Institute of Commerce, 214153, China

Abstract: This paper designs the signal parameter identification methodology for the signal which is composed of the sine components and cosine components. With the help of the gradient search, a stochastic gradient modelling method is presented to estimate all of the trait parameters of the multiple sine-cosine components. Further, some improvement schemes are designed to be aimed at enhancing the precision and convergence speed. Moreover, a rolling optimisation loss function based on the cumulated dynamic measurements is proposed to present a highly efficient and high precision signal modelling methodology. Finally, the algorithm emulation is introduced to confirm the feature of the proposed signal modelling methodologies in improving the accuracy of parameter estimation.

Keywords: signal modelling; parameter estimation; multi-frequency signal; gradient search.

DOI: 10.1504/IJMIC.2024.135536

International Journal of Modelling, Identification and Control, 2024 Vol.44 No.1, pp.14 - 22

Received: 10 Jul 2022
Received in revised form: 28 Aug 2022
Accepted: 05 Sep 2022
Published online: 18 Dec 2023 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article