Anti-Swing and Position Control of Single Wheeled Inverted Pendulum Robot (SWIPR)

Introduction

Wheeled inverted pendulums have been a standard benchmark in control engineering problems (Lim et al. 2011). Single wheeled inverted pendulum robot (SWIPR) comprises of a robot chassis mounted on a single wheel. SWIPR is not only capable of moving in two dimensions but can also perform 360° orientation rotation while maintaining its position. The balance control of SWIPR is widely used in control of Humanoid robots. The stabilization of SWIPR is complicated and involves packaging of all materials in a single wheel (Xu and Au, 2004). Gyrover is one of the well-known single-wheel mobile robot which uses gyroscopic effect for balancing. A lot of research has been done and control methodologies have been proposed in past few decades. Rashid (2007) developed a simulation platform for testing different control techniques to stabilize a single wheeled mobile robot. The graphic representation of the robot, dynamic solution and control schemes were integrated on common computer platform using visual basic. Takagi-Sugeno fuzzy controller was further used for extracting twenty-five fuzzy rules. Huang (2010) derived a model for one-wheeled vehicles (OWVs). The analysis of system stability and controllability were evaluated through simulations. A concise and reliable method through system pole-placement and linear quadratic regulator (LQR) was also proposed to design a self-balancing controller (SBC).

Jae-oh et al. (2011) implemented a unicycle robot by mobile inverted pendulum control method for pitch axis and reaction wheel pendulum control method. The authors assumed that both roll dynamics and pitch dynamics are decoupled. Experimental results proved the validity of proposed technique. Peng et al. (2009) implemented an Omni-directional spherical mobile robot which can move in any direction with no constraints. A fuzzy controller was proposed which can deal with the unknown nonlinearities and external disturbances. The experimental results demonstrated the good performance of the control system. Al-Mamun and Zhu (2010) presented a fuzzy logic controller (FLC) for steering control of single wheel robot. The fuzzy membership functions were optimized using particle swarm optimization (PSO). However, the issue of selecting various functions and parameters for FLC still remained to resolved. Cieslak et al. (2011) presented a design process for a mono-wheel robot. The process includes building a theoretical model, designing a mechanical structure, simulating the design, building a prototype and testing it. The study focuses on the self-stabilization problem encountered in mono-wheel structure and shows the testing results for the case.