UBC Theses and Dissertations
Use of kinematic redundancy for design optimization in robots Wang, Yuyan
For a manipulator mounted on a dynamic structure, as in space applications, the interactions with the robot and its supporting structure can play a crucial role. For example, stability and accurate control of space station operations are critical; therefore, methods that minimize the dynamic interactions (force and torque ) between the space station and robot manipulators are needed. This thesis evaluates two such approaches, and extends these ideas to the design of space robots. General simulations of trajectory optimization schemes are conducted based on the redundancy model proposed by de Silva , , . Furthermore, this optimization theory is extended to cover robot parameter design for a redundant manipulator. Two techniques are developed. First method is the “Joint Angle Manifold Approach”, which uses a path integral to define the global cost function as the sum of the terms contributed by the reaction forces along the robot trajectory, which are described mathematically as different topological manifolds. The final cost function becomes a polynomial of the design parameters i. After calculating the cost function in each manifold, the one with the minimal value was selected for further optimization to search for the optimal link parameter solution. The second approach uses a global time integral to define the cost function as the sum of two terms; one term describing the time integral of the instantaneous base reaction cost function and the second term describing the peak value of all instantaneous cost function curves. The global cost function is then optimized against the design parameters which are link lengths 1, using a numerical optimization method. Both methodologies are simulated for a three degree-of-freedom manipulator.
Item Citations and Data