Reinforcement structure/parameter learning for neural-network-based fuzzy logic control systems

Lin, Chin-Teng; Lee, C.S.G.

doi:10.1109/fuzzy.1993.327458

Cited by 43 publications

(32 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To evaluate the operation of the fuzzy goal coordination system, with the use of reinforcement learning, we define a critic vector [17] and develop a method to train the new coordination strategy. The training is based on minimizing the energy of the critic vector.…”

Section: Reinforcement Learningmentioning

confidence: 99%

A Reinforcement Learning Approach to Intelligent Goal Coordination of Two-Level Large-Scale Control Systems

Sadati¹,

Dumont²

2011

Advances in Reinforcement Learning

View full text Add to dashboard Cite

Section: Reinforcement Learningmentioning

confidence: 99%

A Reinforcement Learning Approach to Intelligent Goal Coordination of Two-Level Large-Scale Control Systems

Sadati¹,

Dumont²

2011

Advances in Reinforcement Learning

View full text Add to dashboard Cite

“…We have termed this improved multilayer perceptron (IMLP) versus the simple multilayer perceptron (SMLP). A great number of neuro-fuzzy topologies and different classes of learning algorithms to adjust the parameters and interconnections between neurons have been proposed in the literature [42]- [44], [31], [40], [45]. The NFN topologies have often been built to reflect the fuzzy inference process in TABLE VI NOISE ANALYSIS USING THE PROPOSED METHOD TABLE VII COMPARISONS WITH ROVATTI'S APPROACH TABLE VIII COMPARISON OF SOFRG WITH OTHER PARADIGMS FOR FUNCTION APPROXIMATION a network structure whose neurons and connections imitate or perform similar operations to those carried out in a fuzzy system.…”

Section: B Comparison Of Sofrg With Other Paradigms For Function Appmentioning

confidence: 99%

“…As representative examples of neural network structures, the multilayer perceptron (MLP) [38], the radial basis function (RBF) [39] and the neuro-fuzzy network (NFN) [40] will be considered.…”

Section: B Comparison Of Sofrg With Other Paradigms For Function Appmentioning

confidence: 99%

Self-organized fuzzy system generation from training examples

Rojas

Pomares

Ortega

et al. 2000

IEEE Trans. Fuzzy Syst.

129

View full text Add to dashboard Cite

Abstract-In the synthesis of a fuzzy system two steps are generally employed: the identification of a structure and the optimization of the parameters defining it. This paper presents a methodology to automatically perform these two steps in conjunction using a three-phase approach to construct a fuzzy system from numerical data. Phase 1 outlines the membership functions and system rules for a specific structure, starting from a very simple initial topology. Phase 2 decides a new and more suitable topology with the information received from the previous step; it determines for which variable the number of fuzzy sets used to discretize the domain must be increased and where these new fuzzy sets should be located. This, in turn, decides in a dynamic way in which part of the input space the number of fuzzy rules should be increased. Phase 3 selects from the different structures obtained to construct a fuzzy system the one providing the best compromise between the accuracy of the approximation and the complexity of the rule set. The accuracy and complexity of the fuzzy system derived by the proposed self-organized fuzzy rule generation procedure (SOFRG) are studied for the problem of function approximation. Simulation results are compared with other methodologies such as artificial neural networks, neuro-fuzzy systems, and genetic algorithms.Index Terms-Function approximation, fuzzy system design, generation of membership functions and rules.

show abstract

“…Controller design based on neuro-fuzzy structure [15] includes both advantages. Controller design based on learning algorithms has been earlier reported [16][17][18][19][20]. A robust neural NARX controller design with evolution and learning is proposed in this paper.…”

Section: Introductionmentioning

confidence: 99%

Instrumentation and control

Roy¹

1982

The Efficient Use of Energy

View full text Add to dashboard Cite

A robust neural NARX controller design using evolution and learning is proposed in this paper. Parameters of neural NARX controller are encoded in individual member (chromosome) of GA population. Neural NARX controller learns stabilizing response at an operating point of plant as best fitness reaches large steady value during successive generations. Stable operating region of neural NARX controller is enlarged by incrementally allowing learn more operating points. Best neural NARX controller in last generation is saved as designed neural NARX controller. Proposed approach is validated on a cart pole nonlinear plant model. Optimal stabilizing response with designed neural NARX controller over wide range of operating points is obtained. KeywordsNon-linear auto regressive with exogenous input (NARX), Reinforcement (RF) signal, Performance index (PI) Genetic algorithms (GAs).

show abstract

Reinforcement structure/parameter learning for neural-network-based fuzzy logic control systems

Cited by 43 publications

References 11 publications

A Reinforcement Learning Approach to Intelligent Goal Coordination of Two-Level Large-Scale Control Systems

A Reinforcement Learning Approach to Intelligent Goal Coordination of Two-Level Large-Scale Control Systems

Self-organized fuzzy system generation from training examples

Instrumentation and control

Contact Info

Product

Resources

About