Singularity-based calibration—a novel approach for absolute-accuracy enhancement of parallel robots

Due to different reasons such as manufacturing and assembly errors, the real geometry of a robot manipulator usually deviates from that defined by the kinematic model which is used for robot control. As a consequence, the absolute accuracy of the robot is limited and offline generated robot programs cannot be executed with sufficient accuracy. With the help of robot calibration, these shortcomings can be overcome. The underlying concept is to estimate the parameters of the kinematic model based on redundant measurements. This data can be used to alter the kinematic model so that it matches the reality as closely as possible. This is ultimately used to correct the parameters in the controller model which thus improves the obtainable absolute accuracy.

This article covers a new and innovative approach for robot calibration which can be applied to parallel robots. In comparison to all known calibration techniques, this novel scheme has the advantage in that it does not rely on any additional calibration hardware. In addition to being cost effective, this method is simple to use as it can be completely automated.