Global calibration scheme using an FCMAC learning algorithm

A global calibration scheme is proposed to resolve the coordinate equivalence problem in integrating CAD (Computer Aided Design) systems and robot manipulators. Current robot calibration schemes are inevitably subject to a certain degree of locality; i.e., the calibrated error parameters (CEPs) will produce the desired accuracy only in certain regions of the robot workspace. This is mainly because of incomplete modeling of errors, resulting in imprecision, and because only a limited amount of measured data is available for identifying the CEPs. To overcome the locality problem, we propose that, first, measurement space analysis be performed to divide the workspace into local regions, and that representative sets of CEPs then be selected from each local region. A learning algorithm based on the FCMAC (Fuzzy Cerebellar Model Articulation Controller) neural network is then employed to generate appropriate sets of CEPs for the whole workspace, based on finite sets of CEPs derived from the measured data. Simulations and experiments that verify the effectiveness of the proposed global scheme are described.