Optimal Actuator Sizing for Robotic Manipulators Based on Local Dynamic Criteria

Abstract

Robotic manipulators provide general, programmable motion paths and force functions to carry out processes of a high level of dexterity and flexibility. These systems are characterized by several degrees of freedom of controllable motion. As a consequence the resulting mechanical structure contains a very large number of design values including geometric, mass, compliance, strength, and prime mover parameters [1]. The analysis on which to base the design methods involves the multivariable mathematical relations between these design parameters and the manipulator’s force and motion states which are extraordinarily complex, nonlinear, and highly coupled. Computer-aided procedures for systems of this class become an imperative in order to establish the dynamic formulation, select rational design specifications, and to evaluate the system’s operating characteristics both locally and globally. This paper suggests some applications of optimization techniques to augment the existing analysis formulation in the literature and to create a more powerful foundation for the design of manipulator structures. This enhanced computational capability is based on position-dependent kinematic and modeling coefficients [6] which explicitly demonstrate the role of significant physical parameters in the design process. Specific examples dealing with optimal distribution of actuator load capacity are given in the paper which improves the system’s load capacity or enhances its speed and acceleration capability within the local neighborhood of a given configuration of the manipulator.