ACTUATOR GAIN DISTRIBUTIONS IN SERIAL ROBOTIC MANIPULATORS TO MEET SPECIFIED TASK REQUIREMENTS

摘要

A serial robotic manipulator arm is a complex electromechanical system whose performance is highly characterized by its actuators. The actuator itself is a complex nonlinear system whose performance can be represented by the speed and torque capabilities of its motor and its accuracy depends on the resolution of the encoder as well as its ability to resist deformations under load. The mechanical gain associated with the transmission is critical to the overall performance of the actuator since it amplifies the motor torque thus improving the force capability of the manipulator housing it, reduces the motor speed to a suitable output speed operating range, enables an improvement in responsiveness (acceleration) and amplifies the stiffness improving the precision under load of the overall system. In this work, a basic analytic process that can be used to manage the actuator gain parameters to obtain an improved arm design based on a set of desired/required performance specifications will be laid out. Key to this analytic process is the mapping of the actuator parameters (speed, torque, stiffness and encoder resolution) to their effective values at the system output via the mechanical gains of the actuators as well as the effective mechanical gains of the manipulator. This forward mapping of the actuator parameters allows the designer to determine how each of the parameters influences the functional capacity of the serial manipulator arm. The actuator gains are then distributed along the effective length of the manipulator to determine the distribution effects on the performance capabilities of the system. The analytic formulation is used to address the issue of configuration management of serial robotic manipulators where the goal is to assemble a system from a finite set of components that meets some required performance specifications. To this end, two examples demonstrating a solution of the configuration management problem are presented. In the first, a manipulator is configured that isrnintended for light-duty applications while in the second, several manipulators intended for medium and heavy-duty applications are configured. The analytic process developed in this work can reduce the effort in the initial phases of the design process and the total number of design iterations can be reduced.