The aim of this paper is to propose a new more efficient self-organizing fuzzy cerebellar model articulation controller (CMAC) for uncertain nonlinear systems. The proposed controller uses a prior Gaussian membership function (GMF) and a present GMF on each layer of CMAC to make a new mixed GMF. Then the inputs can concurrently stir the present state and prior state to regulate the suitable errors so that the new self-organizing fuzzy CMAC (NSOFC) is able to learn the parameters more efficient, reduce the computational loading, anticipate the subsequent state of the inputs, and automatically develop the structure of the NSOFC. The sliding mode control is used to simplify the inputs. By using the self-organizing structure in the NSOFC, the layers will be increased or decreased systematically. The control system consists of an NSOFC and a compensation controller. The NSOFC is the main tracking controller, which imitates an ideal controller, and the compensator expels the approximation error. An adaptive PI method is utilized to online update the parameters for the proposed NSOFC in order to achieve more efficient control performance. The studies of an inverted double pendulum system point out that favorable tracking performance is obtained by using the proposed method.
|Original language||Traditional Chinese|
|Publication status||Published - Nov 22 2019|
Huynh, T. T., Lin, C. M., Le, T-L., Cho, H-Y., Pham, T-T. T., Le, N. Q. K., & Chao, F. (2019). A New Self-Organizing Fuzzy Cerebellar Model Articulation Controller for Uncertain Nonlinear Systems Using Overlapped Gaussian Membership Functions. IEEE Access.