作者: S. Karnalasadan , A.A. Ghandakly
DOI: 10.1109/CIMSA.2004.1397257
关键词:
摘要: This paper presents a novel neural network based intelligent model reference adaptive controller. In this scheme the supervisory loop (ISL) is incorporated into traditional controller (MRAC) framework by utilizing an online growing dynamic radial basis function (RBFNN) structure in parallel with it. The idea to control plant direct MRAC suitable single model, and at same time respond multimodal dynamics on line tuning of RBFNN designed order precisely track system output desired command signal trajectory, regardless multimodality and/or unmodeled dynamics. updating details width, centers weights are derived ensure error reduction for improved tracking accuracy. importance proposed its ability perform effectively even when mode swings without using multiple concept or if can be established. Further, will able trajectory showing performance confirmed applying angular position robotic manipulator under tip load variations.