Partial-Neurons-Based H∞ State Estimation for Time-Varying Neural Networks Subject to Randomly Occurring Time Delays under Variance Constraint

• Published in: Neural processing letters Vol. 55; no. 6; pp. 8285 - 8307
• Main Authors: Hu, Jun, Gao, Yan, Chen, Cai, Du, Junhua, Jia, Chaoqing
• Format: Journal Article
• Language: English
• Published: New York Springer US 01.12.2023; Springer Nature B.V
• Subjects: Algorithms; Artificial Intelligence; Complex Systems; Computational Intelligence; Computer Science; Error analysis; H infinity; Methods; Neural networks; Neurons; Performance indices; Random variables; Recurrent neural networks; State estimation; Variance; Randomly occurring time delays; performance; Time-varying recurrent neural networks; Variance constraint; Partial-neurons-based state estimation
• ISSN: 1370-4621; 1573-773X
• DOI: 10.1007/s11063-023-11312-2

This paper discusses the issue of partial-neurons-based H ∞ state estimation for time-varying recurrent neural networks subject to randomly occurring time delays under variance constraint index. The measurement outputs are allowed to be available only at certain neurons. In addition, a random variable is introduced to model the randomly occurring time delays with certain occurrence probability. The aim is to propose the non-augmented partial-neurons-based state estimation strategy. Accordingly, some sufficient conditions are given to ensure two indices including the pre-determined H ∞ performance index and the error variance boundedness via the stochastic analysis approach. Finally, a simulation example is used to demonstrate the feasibility of presented partial-neurons-based H ∞ state estimation algorithm.