A Deterministic and Probabilistic Approach for Robust Optimal Design of a 6-DOF Haptic Device

Abstract

This work suggests a two-stage approach for robust optimal design of 6-DOF haptic devices based on a sequence of deterministic and probabilistic analyses with a multi-objective genetic algorithm and the Monte-Carlo method. The presented model-based design robust optimization approach consider simultaneously the kinematic, dynamic, and kinetostatic characteristics of the device in both a constant and a dexterous workspace in order to find a set of optimal design parameter values for structural configuration and dimensioning. Design evaluation is carried out based on local and global indices, like workspace volume, quasi-static torque requirements for the actuators, kinematic isotropy, dynamic isotropy, stiffness isotropy, and natural frequencies of the device. These indices were defined based on focused kinematic, dynamic, and stiffness models. A novel procedure to evaluate local indices at a singularity-free point in the dexterous workspace is presented. The deterministic optimization approach neglects the effects from variations of design variables, e.g. due to tolerances. A Monte-Carlo simulation was carried out to obtain the response variation of the design indices when independent design parameters are simultaneously regarded as uncertain variables. It has been observed that numerical evaluation of performance indices depends of the type of workspace used during optimization. To verify the effectiveness of the proposed procedure, the performance indices were evaluated and compared in constant orientation and in dexterous workspace.