Optimal configuration selection method for stiffness identification of serial manipulators based on the κf(A)−1 criterion

In real industrial environment, the stiffness identification accuracy of manipulators is affected by various measurement errors. However, research that deals with the inevitable error perturbation is scarce. The κ F ( A ) −1 criterion is adopted for measurement configuration selection to solve this problem according to the perturbation analysis and derivation of solutions to systems of stiffness identification equations. The optimal measurement configurations are finally obtained using the DETMAX optimization algorithm based on the deduced criterion, which is the main contribution of this work. Results illustrate that the optimal configurations optimized by the DETMAX algorithm based on the κ F ( A ) −1 criterion can better overcome the influence of measurement errors, improve the identification accuracy, and reduce the compensated position error and fluctuation compared with the other optimization algorithms. Furthermore, the proposed criterion can be applied to the stiffness identification of serial manipulators in a real industrial environment.