國立臺灣師範大學電機工程學系Y.-G. LeuW.-C. LeuW.-Y. WangZ.-H. Lee2014-10-302014-10-302010-07-03http://rportal.lib.ntnu.edu.tw/handle/20.500.12235/31994By using B-spline neural networks, an adaptive nonlinear parametric control scheme for nonlinear systems is proposed in this paper. The control scheme which is utilized to design the control input incorporates the adaptive control design technique with the mean-estimation B-spline neural networks. Compared with other neural networks, the B-spline neural networks possess output behavior the characteristic feature of locally controlling. Therefore, they are very suitable to online estimate system dynamics by tuning both control and knot points. The B-spline neural networks with a mean苟stimation technique are used in an attempt to avoid difficulty of differentiating B-spline basis functions. In addition, two robust controllers are used to compensate un modeling dynamics. Finally, an example is provided to demonstrate the feasibility of the proposed scheme, and a comparative study is given by computer simulation.B-spline functionsneural networksadaptive controlnonlinear systems