Information Sciences: an International Journal
Hierarchical cluster-based multispecies particle-swarm optimization for fuzzy-system optimization
IEEE Transactions on Fuzzy Systems
Designing fuzzy-rule-based systems using continuous ant-colony optimization
IEEE Transactions on Fuzzy Systems
Knowledge acquisition in fuzzy-rule-based systems with particle-swarm optimization
IEEE Transactions on Fuzzy Systems
ICIRA'11 Proceedings of the 4th international conference on Intelligent Robotics and Applications - Volume Part II
Robust PI controller design with respect to fuzzy sensitivity margins
Applied Soft Computing
Differential evolution with local information for neuro-fuzzy systems optimisation
Knowledge-Based Systems
International Journal of Intelligent Information and Database Systems
International Journal of Intelligent Information and Database Systems
An observer-based adaptive neural network tracking control of robotic systems
Applied Soft Computing
Hi-index | 0.00 |
This paper proposes a novel method for the incremental design and optimization of first order Tagaki-Sugeno-Kang (TSK) fuzzy controllers by means of an evolutionary algorithm. Starting with a single linear control law, the controller structure is gradually refined during the evolution. Structural augmentation is intertwined with evolutionary adaptation of the additional parameters with the objective not only to improve the control performance but also to maximize the stability region of the nonlinear system. From the viewpoint of optimization the proposed method follows a divide-and-conquer approach. Additional rules and their parameters are introduced into the controller structure in a neutral fashion, such that the adaptations of the less complex controller in the previous stage are initially preserved. The proposed scheme is evaluated at the task of TSK fuzzy controller design for the upswing and stabilization of a rotational inverted pendulum. In the first case, the objective is a time optimal controller that upswings the pendulum in to the upper equilibrium point in shortest time. The stabilizing controller is designed as a state optimal controller. In a second application the optimization method is applied to the design of a fuzzy controller for vision-based mobile robot navigation. The results demonstrate that the incremental scheme generates solutions that are similar in control performance to pure parameter optimization of only the gains of a TSK system. Even more important, whereas direct optimization of control systems with more than 35 rules fails to identify a stabilizing control law, the incremental scheme optimizes fuzzy state-space partitions and gains for hundreds of rules.