An Online GA-Based Output-Feedback Direct Adaptive Fuzzy-Neural Controller for Uncertain Nonlinear Systems

Abstract

In this paper, we propose a novel design of a GA-based output-feedback direct adaptive fuzzy-neural controller (GODAF controller) for uncertain nonlinear dynamical systems. The weighting factors of the direct adaptive fuzzy-neural controller can successfully be tuned online via a GA approach. Because of the capability of genetic algorithms (GAs) in directed random search for global optimization, one is used to evolutionarily obtain the optimal weighting factors for the fuzzy-neural network. Specifically, we use a reduced-form genetic algorithm (RGA) to adjust the weightings of the fuzzy-neural network. In RGA, a sequential-search -based crossover point (SSCP) method determines a suitable crossover point before a single gene crossover actually takes place so that the speed of searching for an optimal weighting vector of the fuzzy-neural network can be improved. A new fitness function for online tuning the weighting vector of the fuzzy-neural controller is established by the Lyapunov design approach. A supervisory controller is incorporated into the GODAF controller to guarantee the stability of the closed-loop nonlinear system. Examples of nonlinear systems controlled by the GODAF controller are demonstrated to illustrate the effectiveness of the proposed method.

Observer-based Variable Universe Adaptive Fuzzy Controller Without Additional Dynamic Order

“…Ωẽ e e = ẽ e e| ẽ e e μ λmin(Q2) (26) whose radius can be made also arbitrarily small by performing similar operations. The parameter errorsβ andε are bounded and will converge to sets Ωβ and Ωε, respectively.…”

Decentralized direct adaptive output feedback fuzzy H ∞ tracking design of large-scale nonaffine nonlinear systems

“…Then, output feedback or observer-based control schemes should be applied to such difficult cases. Recently, many observer-based indirect [18][19][20][21][22] and direct [22][23][24][25][26] adaptive fuzzy or fuzzy-neural control methods have been developed for nonlinear systems as well. However, these control schemes have the following limitations.…”

“…Secondly, that of [23], whose scheme does not utilize strictly positive-realness (SPR), may be impractical as a matter of fact in that the presented feedback and adaptation mechanisms require the observation error dynamical vector to be available online, but the authors have yet to address themselves to the error vector's availability. Thirdly, there exist common drawbacks in the fuzzy-neural control approaches of [24,25] on the grounds that their control laws v cannot ensure the closed-loop systems to be stable in the presence of the filters. Fourthly, the output feedback fuzzy controller suggested in [26], as claimed there, guarantees that the tracking error is only uniformly ultimately bounded (UUB) rather than uniformly asymptotically stable (UAS), which is its inherent limitation.…”

GA-Based Learning Algorithms to Identify Fuzzy Rules for Fuzzy Neural Networks

“…Several authors have proposed a genetic algorithm for fuzzy neural parameters optimization to adjust the control points of membership functions or to tune the weightings [9][10][11][12][13][14]. The pioneer was Karr [9] , who used GAs to adjust membership functions.…”

“…He has used GA to tune membership functions at the precondition part of fuzzy rules, while the least-squares estimate method has been used to tune parameters at the consequent part. Wang et al [12] have proposed GA-based approach for a feedback direct adaptive fuzzy-neural controller to tune the online weighting factors. Specifically, they have used a reduced-form genetic algorithm (RGA) to adjust the weightings of the fuzzy-neural network.…”