On-Line Neuro Identification of Uncertain Systems Based on Scaling and Explicit Feedback

Vargas, José A. R.; Gularte, Kevin H. M.; Hemerly, Elder Moreira

doi:10.1007/s40313-013-0082-6

Cited by 14 publications

(7 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Neural identification models commonly employed are the linearly and nonlinearly parameterized, which can be by nature static or dynamic. Their weights are often adjusted using gradientbased schemes, as the backpropagation algorithm, or their robust modifications [2,19,20,21,10,66,67,68,69,70]. The most widely-used robust modifications in neuro-identification are the , switching-, " 1 , parameter projection, and dead zone [1][2][3][4][5][6][7][8][9][10], which avoid the parameter drift.…”

Section: State Of the Art Review Of Identification Based On Single-hidden Layer Neural Networkmentioning

confidence: 99%

“…For instance, in [67], the neuro-identification of a general class of uncertain continuous-time dynamical systems was proposed, and a -modification adaptive law for the weights of recurrent high-order neural networks (RHONNs) was chosen to ensure that the state error converges to the neighborhood of zero. More recently, in [20,21,66], neuro identification schemes for open loop systems were proposed. In [20,21] was established the conditions to ensure the asymptotical convergence of the residual state error to zero, even in the presence of approximation error and bounded internal or external perturbations.…”

Section: State Of the Art Review Of Identification Based On Single-hidden Layer Neural Networkmentioning

confidence: 99%

“…The convergence of the state error to zero in both works ( [20,21]) was based, among other, on the previous knowledge of bounds for the approximation error and perturbations, which are usually unknown in practice. In [66], an identification scheme based on a dynamical neural model with scaling and a robust weight adaptive law was proposed. The main peculiarity of [66] is that the residual state error is directly related to two design matrices, which allow the residual state error to be arbitrarily and easily reduced.…”

Section: State Of the Art Review Of Identification Based On Single-hidden Layer Neural Networkmentioning

confidence: 99%

“…In [66], an identification scheme based on a dynamical neural model with scaling and a robust weight adaptive law was proposed. The main peculiarity of [66] is that the residual state error is directly related to two design matrices, which allow the residual state error to be arbitrarily and easily reduced.…”

Section: State Of the Art Review Of Identification Based On Single-hidden Layer Neural Networkmentioning

confidence: 99%

“…Despite the remarkable theoretical contribution in these works ( [67,20,21,66]), they are all based on linearly parameterized neural networks and consequently, in general, suffer from "the curse of dimensionality". That is, these models have a poor capability of interpolation and require a large number of basic functions to deal with multi-dimensional inputs.…”

Section: State Of the Art Review Of Identification Based On Single-hidden Layer Neural Networkmentioning

confidence: 99%

See 4 more Smart Citations

Identification of nonlinear systems based on extreme learning machine

Grzeidak¹

View full text Add to dashboard Cite

O resultado das simulações para o sistema caótico unificado mostram a eficácia e o desempenho da abordagem proposta na presença de distúrbios. Adicionalmente, para mostrar a aplicabilidade do algoritmo de aprendizado proposto para sistemas com várias dimensões, uma simulação com um sistema hipercaótico complexo é exibida. Finalmente, uma comparação do algoritmo proposto com [2] é realizada para mostrar as vantagens e peculiaridades do método proposto na presença de distúrbios. Capítulo 6 resume as contribuições teóricas da pesquisa bem como os resultados obtidos. Sugestões para pesquisa futura também são discutidas. Os apêndices contém a implementação via software dos identificadores neurais propostos nesta dissertação.3 Technical Background .

show abstract

Section: State Of the Art Review Of Identification Based On Single-hidden Layer Neural Networkmentioning

confidence: 99%

Section: State Of the Art Review Of Identification Based On Single-hidden Layer Neural Networkmentioning

confidence: 99%

Section: State Of the Art Review Of Identification Based On Single-hidden Layer Neural Networkmentioning

confidence: 99%

Section: State Of the Art Review Of Identification Based On Single-hidden Layer Neural Networkmentioning

confidence: 99%

Section: State Of the Art Review Of Identification Based On Single-hidden Layer Neural Networkmentioning

confidence: 99%

See 3 more Smart Citations

Identification of nonlinear systems based on extreme learning machine

Grzeidak¹

View full text Add to dashboard Cite

show abstract

Robust extreme learning machine neural approach for uncertain nonlinear hyper‐chaotic system identification

Ánh

Kien

2021

Intl J Robust & Nonlinear

View full text Add to dashboard Cite

This paper proposes a novel nonlinearly parameterized advanced single-hidden layer neural extreme learning machine (ASHLN-ELM) model in which the hidden and output weighting values are simultaneously updated using adaptively robust rules that are implemented based on Lyapunov stability principle. The proposed scheme guarantees the fast convergence speed of the state-estimation residual errors bounded to null regarding to the influence of time-varied disturbances. Additionally, proposed method needs no any knowledge related to desired weighting values or required approximating error. Typical uncertain hyper-chaotic benchmark systems are used as to verify the new ASHLN-ELM approach and to demonstrate the efficiency and the robustness of proposed method. K E Y W O R D Sadvanced single hidden layer neural extreme learning machine (ASHLN-ELM) model, extreme learning machine (ELM) algorithm, Lyapunov stability principle, online adaptive identification rule, uncertain nonlinear hyper-chaotic system identification

show abstract