Optimal architecture design of parallel manipulators for best accuracy

Presents a design optimization method of parallel manipulators for best accuracy. An accurate kinematic error model that relates every structural error source in the manipulator's structure to end-effector pose errors is derived for a HexaSlide type parallel manipulator. Based on the error model, a measure of accuracy, error amplification factor, is introduced for an optimum design formulation with constraints on workspace and design variable limits. Then, design optimization for best accuracy has been performed by using a nonlinear optimization technique. Optimization results have been validated by Monte Carlo statistical simulation technique. The optimized design shows smaller error than the initial design.