Indirect adaptive nonlinear self-balancing and station keeping for omnidirectional riding chair

This paper presents indirect adaptive self-balancing and station keeping control methods using recurrent Wavelet Fuzzy CMAC (RWFCMAC) for an omnidirectional ball-driven chair in presence of significant system uncertainties. By backstepping, sliding-mode control and RWFCMAC, the self-balancing controller is synthesized to follow the rider´s inclination angles in both two axes (x-z and y-z axis), and the station-keeping controller is designed to allow the rider to maintain the vehicle at the same place. The RWFCMAC is designed to online learning the uncertainties caused by riders´ weights and different unknown frictions between the ball and terrain surfaces. The superior performance and merit of the proposed control methods are well exemplified by comparing to two existing controllers.