Investigation on the impact of nongeometric uncertainty in dynamic performance of serial and parallel robot manipulators

摘要

Uncertainty in robot manipulations is a fundamental source of inaccuracy. Most nongeometric uncertainties is caused by the mechanical components of robot manipulators such as gear transmission systems and joints. This study integrates existing uncertainty models for practical applications of both three-degrees-of-freedom serial robot and parallel robot (closed-loop 5R mechanism). We first review the literature on uncertainty and existing methods for improving the precision of robot manipulation. For existing methods for uncertainty modeling: (1) the mathematical model of a harmonic drive considering kinematic error and compliance is derived; (2) joint clearance analysis that includes models of contact force and friction force is presented; (3) Archard's wear model is used to estimate the effect of joint wear, based on which the geometry of joints is updated. Computer simulation is then used to demonstrate the application of the proposed method for serial and parallel robot manipulators. Simulations are used to examine the behavior of robot manipulators operating under uncertainty, with performance-evaluated positional error and trajectory error. The proposed method can provide some insight into nongeometric uncertainty in robot manipulations. The proposed methodology can be further used to improve the accuracy of robot manipulation under uncertainty by identifying uncertainty parameters.