Online Adaptive Control of Non-linear Plants Using Neural Networks with Application to Temperature Control System

Teshnehlab

et al. 2015

JAIDM

The stability of learning rate in neural network identifiers and controllers is one of the challenging issues, which attract many researchers' interest in neural networks. This paper suggests adaptive gradient descent algorithm with stable learning laws for modified dynamic neural network (MDNN) and studies the stability of this algorithm. Also, stable learning algorithm for parameters of MDNN is proposed. By the proposed method, some constraints are obtained for learning rate. Lyapunov stability theory is applied to study the stability of the proposed algorithm. The Lyapunov stability theory guaranteed the stability of the learning algorithm. In the proposed method, the learning rate can be calculated online and will provide an adaptive learning rate for the MDNN structure. Simulation results are given to validate the results.

Section: Introductionmentioning

confidence: 99%

Designing stable neural identifier based on Lyapunov method

Teshnehlab

et al. 2015

JAIDM

2015 IEEE International Conference on Industrial Technology (ICIT)

“…It has been shown that neural networks based inverse dynamics control is applied to both linear and nonlinear plants [18][19][20]. Numerous applications of neural adaptive tracking can be found in robotics, industrial and process control [21][22][23][24][25][26][27][28][29][30]. Many researchers have also presented control techniques formed by combining adaptation capabilities of neural networks with conventional nonlinear control methodologies.…”

Section: Introductionmentioning

confidence: 99%

Lyapunov stability criterion based neural inverse tracking for unknown dynamic plants

Aftab

Shafiq

Yousef

2015

This paper presents a Lyapunov function based neural network tracking control strategy for single-input-singleoutput nonlinear dynamic systems. The proposed architecture is composed of two feed-forward neural networks operating as controller and estimator in a unified framework. The network parameters are tuned online with a Lyapunov function based backpropagation learning algorithm. The closed-loop error convergence and stability are analyzed with Lyapunov stability theory. Two simulation case studies are included that successfully validate the proposed controller performance. Keywords-Lyapunov function; neural inverse tracking; stable adaptive tracking; direct adaptive inverse control; indirect adaptive inverse controlI.

“…Most of the parametric adaptive control scheme involves a very complex mathematical derivation [1][2][3]. Thus, the design process becomes difficult and almost impossible for a high order plant.…”

Section: Introductionmentioning

confidence: 99%

Adaptive Neuro-Controller Based on Hybrid Multi Layered Perceptron Network for Dynamic Systems

Sharun¹,

Mashor²,

Jaafar³

et al. 2012

CONTROL

In this paper, an intelligent controller namely Adaptive Neuro-controller (ANC) based on Hybrid Multi Layered Perceptron (HMLP) network has been developed for dynamic systems. The performance of ANC has been compared with the ANC based on Multi Layered Perceptron (MLP) network and Adaptive Parametric Black Box (APBB) Controller. The comparison are based on the time response and the capability of the controlled output to track the model reference output. All controllers are based on a black box approach that offers simpler design approach. The Model Reference Adaptive System (MRAS) has been used to generate the desired output path and to ensure the output of the controlled system follows the output of the reference model. Weighted Recursive Least Square (WRLS) algorithm has been used to adjust the controller parameters in order to minimize the error between the plant output and the model reference output. The controllers have been tested using a linear plant and a nonlinear plant with several varying operating conditions such as varying gain, noise and disturbance. Based on the simulation results and performance analysis for all controllers, it is observed that ANC based on HMLP network is controllable and more stable than ANC based on MLP network and APBB controller. It is also can be signify that the ANC based on HMLP network is sufficient to control the plants with unpredictable conditions.