Progress in on-line adaptive, learning and evolutionary strategies for fuzzy logic control

In this paper, the eight kinds of on-line adaptive, learning and evolutionary strategies for fuzzy logic control are systematically introduced. All these afore-mentioned strategies have some drawbacks in terms of generalization and formulation. Hence a systematic way of combination and hybridization of these strategies will be very useful for improving the learning capacity and performance of algorithms based on these strategies. It is concluded that the orientation of deep-going pathfinding in the generation and modification of fuzzy control rules or models which is principally based on neural networks combined with genetic algorithms or other algorithms should be able to compensate for the disadvantages of neural networks learning.