Varying‐parameter finite‐time zeroing neural network for solving linear algebraic systems. Issue 16 (1st August 2020)

• Record Type: Journal Article
• Authors: Gerontitis, Dimitrios; Moysis, L.; Stanimirović, Predrag; Katsikis, Vasilios N.; Volos, C.
• Is Part Of: Electronics letters. Volume 56:Issue 16(2020)
• Page Start: 810
• Page End: 813
• Publication Date: 2020-08-01
• Subjects: mathematics computing -- linear algebra -- recurrent neural nets; parameter finite‐time -- linear algebraic systems -- recurrent neural network -- finite‐time convergence -- exponential term -- Zhang neural network dynamical system; Electronics -- Periodicals; 621.381
• DOI: 10.1049/el.2019.4099
• Languages: English
• ISSNs: 0013-5194

Abstract

Abstract : A new recurrent neural network is presented for solving linear algebraic systems with finite‐time convergence. The proposed model includes an exponential term in the Zhang neural network dynamical system, which leads to a faster convergence of the error‐monitoring function in comparison to previous methods. Theoretical analysis, as well as simulation results, validate the efficacy of the proposed model.