ADAPTIVE CONTROLLER DESIGN AND SELF-BALANCING TWO-WHEELED TRANSPORTER (SBTWT

Abstract

With tremendously increased use of the personal transporter vehicle in comparison of (public transporters), the problems like damage caused by carbon dioxide and other greenhouse gas emissions, and environmental and political forces have de-stabilized the global petroleum supply and fuel price hike are growing day by day. So the personal transporter like SegwayTM is an effective solution, the Segway PT can drastically reduce dependence on foreign oil, pollution, greenhouse gas emissions and substantially increase energy efficiency by replacing short-distance single-occupancy car journeys. Its compact yet robust design makes it suitable for a variety of everyday uses and commercial applications, allowing riders to cover distances which would have previously required the use of a traditional vehicle for the purpose of travelling some miles. Such Self-balancing Two-Wheeled Transporters (SBTWT) which are less expensive in comparison to Segway PT can be built using low-tech, off-the-shelf inexpensive components so that such type of vehicle can be easily used like traditional bicycle or scooter to serve the purpose of human transportation. The controlling of these low cost vehicles in such a way so that they offer sturdy capability in rugged off-sidewalk terrains such as trails, bike paths or beachfronts is an open research problem. This work contributes to design the controller for self-balancing two-wheeled transporter so that rider can experience the comfortable standing posture and the motion control in riding. In this work firstly a tracking and controlling of the SBTWT using feedback linearization technique is discussed to achieve the self-balancing and the yaw motion control of the vehicle and the performance is examined with system uncertainty, parameter variation and for different riders and comparative study with existing control techniques. Since the self-balancing two-wheeled transporter is an application of the inverted pendulum, an Adaptive Neuro-Fuzzy Inference Structure (ANFIS) controller is designed for Rotary Inverted Pendulum (RIP). The controller and the inverted pendulum are simulated in the Matlab Simulink environment with the help of ANFIS editor GUI and the performance of this controller is shown in comparison to conventional PID and fuzzy logic controller. The problem with the existing controllers is that the control signal is dependent upon the mathematical modeling of the vehicle. An indirect adaptive controller is designed for SBTWT. Since the control signal is not depend upon the mathematical dynamic equation of the system, the performance of the proposed controller is not affected from environment changes, system parameters changes, uncertainty and disturbances.