A MEBML-based adaptive fuzzy logic controller
Keming Xie, Chanhua Mou, Gang Xie · 2002
In this paper, a new adaptive fuzzy logic controller with online tuning the scaling factor is proposed. By using the information from the fuzzy logic controller and experience rules, the output scaling factor and transforming functions from the fuzzy universal discourse to the basic one in the fuzzy logic controller, are decided. In this way, the controller possesses an adaptive ability. Furthermore, a new evolutionary computing method, called the mind-evolutionary-based machine learning (MEBML), is adopted in this paper. MEBML inherits "colony" and "evolution" of the evolutionism. It jumps the traces of the gene and solves successfully the encoding problem of the genetic algorithm. Simulation illustrates that this new adaptive fizzy controller not only can self-tune the parameters of the controllers online and increase control system qualities, but its algorithm is also simple and easy to be established.