Dynamical Neural Networks that Ensure Exponential Identification Error Convergence

Kosmatopoulos, E.B.; Christodoulou, Manolis A.; Ioannou, Pétros

doi:10.1016/s0893-6080(96)00060-3

Cited by 112 publications

(54 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…A high order neural network is a NN which not only a linear combination of the components, but also of their products are considered, as it is explained in [9]. The RHONN model is flexible and allows incorporating to the neural model a priory information about the system structure with less units.…”

Section: Discrete-time Recurrent High Order Neural Networkmentioning

confidence: 99%

“…It is used the EKF training algorithm and is performed on-line using a series-parallel configuration [17], due to the fact that this configuration constitutes a well approximation method of the real plant by the neural identifier and improve the learning convergence. In this paper, the associated state Q i (k) and measurement R i (k) noises discrete-time covariance matrices are computed as (8)(9) as it is explained in subsection 2.2. All the NN states are initialized in a random way as well as the weights vectors.…”

Section: Furuta Pendulum Identificationmentioning

confidence: 99%

“…On the other hand, artificial neural networks (NN) have become an attractive tool which can be used to construct a model of complex nonlinear processes ( [4], [9], [18]); particularly, the used of discrete-time high order neural networks for identification and control has recently increased ( [2], [4]) where the extended Kalman filter (EKF) based algorithms have been introduced to train these NN and to improve the learning convergence ( [1], [7], [19], [20]), among others.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Real-Time Control Based On High Order Neural Networks Using Stochastic Estimation

Castaneda¹,

Hermosillo²,

Esquivel³

et al. 2016

Anais Do 10. Congresso Brasileiro De Inteligência Computacional

View full text Add to dashboard Cite

Section: Discrete-time Recurrent High Order Neural Networkmentioning

confidence: 99%

Section: Furuta Pendulum Identificationmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Real-Time Control Based On High Order Neural Networks Using Stochastic Estimation

Castaneda¹,

Hermosillo²,

Esquivel³

et al. 2016

Anais Do 10. Congresso Brasileiro De Inteligência Computacional

View full text Add to dashboard Cite

“…Different from the the area of NN-based control where NNs are only functioning in achieving accurate state estimation to pursue the control goal, NN-based system identification further requires the NN weights to converge to their correct values so as to capture the best approximation of the underlying system dynamics [2]. To achieve this objective, the persistent excitation (PE) condition is normally required to be satisfied [5][6][7]. Nevertheless, the PE condition is very difficult to be characterized and usually cannot be verified a priori, especially for identification of nonlinear systems [6].…”

Section: Introductionmentioning

confidence: 99%

Design and performance analysis of deterministic learning of sampled-data nonlinear systems

Yuan

2013

Sci. China Inf. Sci.

View full text Add to dashboard Cite

In this paper, we extend the deterministic learning theory to sampled-data nonlinear systems. Based on the Euler approximate model, the adaptive neural network identifier with a normalized learning algorithm is proposed. It is proven that by properly setting the sampling period, the overall system can be guaranteed to be stable and partial neural network weights can exponentially converge to their optimal values under the satisfaction of the partial persistent excitation (PE) condition. Consequently, locally accurate learning of the nonlinear dynamics can be achieved, and the knowledge can be represented by using constant-weight neural networks. Furthermore, we present a performance analysis for the learning algorithm by developing explicit bounds on the learning rate and accuracy. Several factors that influence learning, including the PE level, the learning gain, and the sampling period, are investigated. Simulation studies are included to demonstrate the effectiveness of the approach.

show abstract

“…In this paper, we use an online identification methodology proposed in [8,9]. The proposed control scheme is based on neural networks and sliding-mode control [10].…”

Section: Introductionmentioning

confidence: 99%

Force tracking neural block control for an electro‐hydraulic actuator via second‐order sliding mode

Loukianov

Sánchez

Lizalde

2007

Intl J Robust & Nonlinear

View full text Add to dashboard Cite

SUMMARYThis paper presents a novel scheme for identification and control of an electro-hydraulic system using recurrent neural networks. The proposed neural network has the nonlinear block control form structure. A sliding-mode control technique is applied then to design a discontinuous controller, which is able to track a force reference trajectory. Due to the presence of an unmodelled dynamics, the standard sliding-mode controller produces oscillations (or 'chattering') in the closed-loop system. The second-order sliding mode is used to eliminate the undesired chattering effect. Simulations are presented to illustrate the results.

show abstract

Dynamical Neural Networks that Ensure Exponential Identification Error Convergence

Cited by 112 publications

References 15 publications

Real-Time Control Based On High Order Neural Networks Using Stochastic Estimation

Real-Time Control Based On High Order Neural Networks Using Stochastic Estimation

Design and performance analysis of deterministic learning of sampled-data nonlinear systems

Force tracking neural block control for an electro‐hydraulic actuator via second‐order sliding mode

Contact Info

Product

Resources

About