An adaptive system for enhancing vehicle bodies assembly using range sensing.

In many of today's industrial robot applications, motions of the manipulator are planned using a prototype workpiece and played back for the execution of the actual manufacturing processes. This thesis addresses the problem of programming and controlling of industrial robots with the aim of responding to the demands for increased reliability and productivity with the desired quality and adaptability to environmental changes. The developed system has been applied to the assembly of automotive Body-In-White parts, namely, the hinge and fender. Robotic systems consist primarily of a manipulator and actuators that drive the manipulator links. Currently available robot controllers require exact knowledge of both robot and motor dynamics and acceleration feedback. Chapters 4, 5 and 6 presents a new approach that does not require exact knowledge of robot-actuator system. In chapter 4, This is achieved by adopting adaptive feedback linearization techniques. Enhancing those techniques, to be applied in an optimal sense by exploiting the Hamilton-Jacobi equation, minimizes the required control effort. The results demonstrate an asymptotic tracking response for the output; this is achieved using minimum control effort. By implementing an observer that guarantees asymptotic stability for the estimation errors, the acceleration feedback constraint is removed, which has a significant impact from a practical point of view since acceleration feedback is not available for most industrial controllers. Moreover, applying this technique removes linear growth constraints on the nonlinearities inherent in the system in order to guarantee global stability. Chapter 5 presents a different approach to the same problem by adopting a robust adaptive motion controller that requires only position measurements. The global stability of the proposed controller tracking performance is proved in the Lyapunov sense. In addition, simulation results are presented to demonstrate the asymptotic tracking performance of the closed-loop system. The significance of the above mentioned technique is that it does not require that a general expression and bound for the control input signal be found. In addition, it does not assume the availability of velocity measurements, which, for practical purposes may not be readily available. In Chapter 6, an adaptive controller ensuring zero force and tracking errors has been proposed. The stability of the proposed controller tracking performance is proven in the Lyapunov sense. The presented method establishes global as opposed to local convergence, in the presence of unmodelled dynamics. (Abstract shortened by UMI.)Dept. of Industrial and Manufacturing Systems Engineering. Paper copy at Leddy Library: Theses & Major Papers - Basement, West Bldg. / Call Number: Thesis1999 .E43. Source: Dissertation Abstracts International, Volume: 61-09, Section: B, page: 4916. Adviser: Waguih El Maraghy. Thesis (Ph.D.)--University of Windsor (Canada), 1999.