Pilot Control and Fault Diagnosis of an Auto-balancing Two-wheeled Cart

博士 === 國立成功大學 === 機械工程學系碩博士班 === 96 === Since the beginning of the new millennium, the auto-balancing two-wheeled cart (ABTWC) has become more and more popular due to its responsive yet precise movement and pollution-free. This dissertation is devoted to investigating both the pilot control and faul...

Full description

Bibliographic Details
Main Authors: Jia-Sheng Hu, 胡家勝
Other Authors: Mi-Ching Tsai
Format: Others
Language:en_US
Published: 2008
Online Access:http://ndltd.ncl.edu.tw/handle/21772937250461080327
Description
Summary:博士 === 國立成功大學 === 機械工程學系碩博士班 === 96 === Since the beginning of the new millennium, the auto-balancing two-wheeled cart (ABTWC) has become more and more popular due to its responsive yet precise movement and pollution-free. This dissertation is devoted to investigating both the pilot control and fault diagnosis technologies for an ABTWC which is inherently unstable and has a non-minimum phase. In this dissertation, we present a pilot control algorithm, which converts two joysticks’ commands into two torque directives for movement. This technique allows the user to operate the ABTWC linearly for both motion and orientation control via a joystick. Also, the dynamics modeling and corresponding system phenomena of the ABTWC are discussed. Since a human being is involved in the operation of an ABTWC, the rider faces the danger of being injured in a fall if any system failure occurs. Therefore, the rider should be warned immediately when any system failure develops, ensuring that proper action can be taken to avoid a dangerous accident. The sensor fault-detection technology for the ABTWC is proposed in this dissertation. A model-based fault-detection filter is designed to detect sensor faults. The observer gain obtained by solving an algebraic Riccati equation in the normalized coprime factorization approach offers some design convenience associated with the fault diagnosis filter. The actuator fault-detection is also investigated. This dissertation employs a PI observer to detect abnormal information in an ABTWC caused by actuator faults and steering load-torques. In order to promptly alert the rider for safety purposes in the event of a malfunction, the decision-making process to identify a critical failure is investigated. A statistical threshold that has the benefits of improving decision-making reliability is investigated for diagnosing a possible abnormal operation and/or a serious system malfunction. The experimental results substantiate that the proposed pilot control and fault diagnosis strategies have the ability, in practice, to improve the ABTWC’s mobility and safety.