Thesis (M.S.)--Ohio University, March, 1992. / Title from PDF t.p.

This study presents a new analytical controller design strategy for the teleoperation of mechanical manipulators aboard the U.S. space station. This controller design strategy emphasizes on the stability of a closed-loop control system involving time delay. Simplified dynamic equations of the Stanford arm are considered as the manipulator model. A local linearizing and decoupling control algorithm is applied to linearize and decouple the dynamic equations. Once the linear form of the manipulator is obtained, a model prediction control loop is constructed and implemented as a digital controller to provide the predictive states information, and a particular model reduction method is applied to yield a reduced-order digital controller. This reduced-order digital controller is a highly self-tuned controller which can control the closed-loop system with time delay by following a specified performance.

No description available.

Control of high-speed, light-weight robotic manipulators is a challenge because of their special dynamic characteristics. In this work, a two-stage control algorithm for the position control of flexible manipulators is proposed. First, the more complex, flexible robot system is replaced by a simplified hypothetical rigid body system (HRRA) with off-line trajectory planning. This reduces the complexity of the controller design for the flexible robotic arm. A parameter-optimization approach was adopted to minimize the difference between these two models in this stage. Also, a comparison of computational efficiency is made among the methods of calculus-of-variations, dynamic-programming, and the proposed parameter-optimization. At the second stage, simple linear state feedback controllers, based on the simplified hypothetical rigid body model, are proposed to control the actual robotic system. With the feedback gains selected properly by the pole-placement and linear quadratic methods, the results show satisfactory achievement of the motion objectives. The algorithm is implemented for a two-link rigid/flexible robotic arm, and the results indicate that the procedure is capable of providing effective control with much simpler computational requirements than those of procedures published previously. / Graduation date: 1992

No description available.

No description available.

No description available.

No description available.

No description available.

Graduate