UBC Theses and Dissertations
UBC Theses and Dissertations
Structural failure identification and control of the multi-module deployable manipulator system (MDMS) Wong, Ho Keung Kenneth
This thesis focuses on the study of dynamics and control, and failure detection and identification (FDI) for the multi-module deployable manipulator system (MDMS), which has been designed and developed in our laboratory. The interest in this unique class of manipulator systems originated mainly from Canada’s contribution to the International Space Station (ISS). In contrast to all the existing space-based robotic manipulators, which have revolute (slew) joints only, the MDMS has a combined prismatic (deployable, telescopic) joint and a revolute joint in each module, and several such modules connected in series to form the manipulator system. This novel design possesses several attractive features in dynamics and control; for an example, reduced number of singular configurations, reduced dynamic interactions, and improved obstacle-avoidance capability for a specified number of degrees of freedom (joints). Research has been performed in the dynamics and control of this class of manipulators, particularly by our laboratory, which has resulted in the design and development of a prototype MDMS. The research presented in this thesis extends this founding work to explore possible types of structural failure in the MDMS and their accurate identification, and the use of kinematic redundancy and controllability to adapt to such failures. Structural failures of a robotic system are critical in remote and hazardous environments such as space and nuclear active sites. If a joint of the manipulator fails in the course of a manoeuvre, it is immensely desirable to be able to complete the specified task automatically without immediate human intervention. This capability will result in higher system reliability, increased safety, better maintainability, improved survivability, and greater cost effectiveness. In the present work, methodology is developed for accurate identification of structural failures in the MDMS, which through the use of a decision making strategy, effective control and kinematic redundancy is capable of satisfactorily executing the intended robotic task even in the presence of structural failure. The method of failure identification is based on Bayes hypothesis testing. First, a possible set of failure modes is defined, and a hypothesis is associated with each considered failure mode. The most likely hypothesis is selected depending on the observations of the response of the manipulator and a suitable test. This test minimizes the maximum risk of accepting a false hypothesis, and accordingly the identification methodology is considered optimal. The implementation of this approach to the detection of structural failures in MDMS is presented in detail. In the developed methodology, a bank of discrete Kalman filters is used for the computation of the hypothesis-conditioned information about the MDMS, which is required in the decision logic. The developed failure identification methodology is rather general, and any structural failure mode can be included in the detection strategy. For the purpose of illustration, three specific joint failure modes are considered: sensor failure, locked joint, and freewheeling or free-sliding joint. Successful identification of these joint failure modes using the developed approach is demonstrated by implementation on a single- and a two-module version of the MDMS. Furthermore, the failure detection and identification (FDA) scheme has been incorporated into the control system for safe operation under failures. Computer simulations are used to demonstrate how a two-module redundant deployable manipulator is able to successfully complete a targeting and pointing task in the presence of a failure where the shoulder joint of the manipulator is locked. The present study provides a good foundation for further studies in fault tolerance or fault recovery for this class of novel robotic manipulators.
Item Citations and Data