Optimal Selection of Measurement Configurations for Robot Calibration Using Modified Simulated Annealing Algorithm

Abstract

The selection of measurement configurations in robot calibration was investigated. The goal was to select a set of robot measurement configurations with modified simulated annealing algorithm so that the effect of measurement and modeling errors in parameter estimation could be minimized. To accelerate the convergence rate, a suitable cooling schedule was devised. The interactive influences of the parameters in cooling schedule on algorithm performance were discussed. Experimental results, comparing with those of the local search algorithm, demonstrate the feasibility of the proposed approach.