A training rule which guarantees finite-region stability for a class of closed-loop neural-network control systems

Kuntanapreeda, Suwat; Fullmer, R. Rees

doi:10.1109/72.501730

Cited by 19 publications

(3 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In [9] and [10] Lewis et al applied a neural network accompanied with a PD controller to a robot model. Their proposed neural network cancels out the nonlinear part of the robot model.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

A Stable Neural Network Controller for Linear Saturated Systems Containing Hysteresis

Rouhani¹,

Menhaj²

2002

Intelligent Automation & Soft Computing

View full text Add to dashboard Cite

This paper introduces a design procedure for determination of weight matrices of an MLP-based controller to locally stabilize a class of multi-input nonlinear systems consisting of a linear system together with a nonlinear saturated actuator with or without hysteresis. No assumption is made on the stability of the open-loop system. Two different design problems are considered: 1) designing a neural network which maximizes the guaranteed domain of stability of the overall system when the actuator bounds are given, and 2) designing a neural network which guarantees the stability of the overall system in a pre-specified region in state-space with a minimum actuator bound. The proposed controller has excellent advantage over the classical ones, regarding its guaranteed domain of stability. The design procedure replaces earlier recursive neural network learning algorithms with a simple optimization problem.

show abstract

“…In [9] and [10] Lewis et al applied a neural network accompanied with a PD controller to a robot model. Their proposed neural network cancels out the nonlinear part of the robot model.…”

Section: Introductionmentioning

confidence: 99%

“…For Riccati equation(10) to have a unique solution, we require that the controllable. This can be achieved only if the linear system is controllable and in addition we must choose 2 w (as follows) controllable.…”

mentioning

confidence: 99%

A Stable Neural Network Controller for Linear Saturated Systems Containing Hysteresis

Rouhani¹,

Menhaj²

2002

Intelligent Automation & Soft Computing

View full text Add to dashboard Cite

show abstract

“…Although this design process is systematic, the global closed-loop system stability may not be guaranteed and the training process is quite time consuming. Stability conditions of the neural-network and neural-fuzzy control systems can be found in [29], [30], and [57]. A fuzzy PID controller was also proposed to control a plant based on the model-free approach.…”

Section: Introductionmentioning

confidence: 99%

Stable and robust fuzzy control for uncertain nonlinear systems

Lam

Leung

Tam

2000

IEEE Trans. Syst., Man, Cybern. A

View full text Add to dashboard Cite

Abstract-This paper presents the stability and robustness analysis for multivariable fuzzy control systems subject to parameter uncertainties based on a single-grid-point (SGP) approach. To perform the analysis, we represent a multivariable nonlinear system using a TS-fuzzy plant model. Three design approaches of fuzzy controllers are introduced to close the feedback loop. By estimating the matrix measures of the system parameters and parameter uncertainties, stability and robustness conditions for different cases are derived. Application examples will be given to show the design procedures and the merits of the proposed fuzzy controller.Index Terms-fuzzy control, nonlinear system robustness, parameter uncertainty, stability.

show abstract