Combined Structural and Dimensional Synthesis of Task Redundant Robot Manipulators

点击翻译按钮获取中文摘要

Research activities on the combined structural and dimensional synthesis so far only have dealt with task-specific robots without redundancy. The aim of this project is to investigate modelling and selection methods in order to offer more alternatives for robots deployed in industry, to enable efficiency increases and to open up fields of application such as medical technology on a larger scale by taking into account field-specific conditions. The project focuses on tasks with five defined degrees of freedom (DoF), e.g. drilling, milling, welding or laser cutting, where rotational symmetry leads to one degree of redundancy of the end effector DoF (task redundancy).The presence of more DoF offers both disadvantages (more actuators, thus higher energy consumption, mass, costs) and advantages (redundancy increases the workspace and helps avoiding singularities). The automated balancing of these advantages and disadvantages and the comparison with existing structures perspectively leads to the use of redundant structures, if they are advantageous for the respective task. As a result, less oversized standard robots are used, which both have structurally more end-effector DoF than necessary for the task (task redundancy) and are oversized in their dimensions. Although these robots fulfill kinematic requirements such as a collision- and singularity-free workspace, their energy consumption is higher than necessary due to the larger masses and additional drives.The aim of the proposal is to extend the methods for redundancy resolution for the combined structural and dimensional synthesis of task-specific robots. This enables task-redundant kinematics whose dimensions are optimized for a given task. This results in robots that are task specific in their dimensions, but not – due to redundancy – in their structure. The task redundancy of the structures to be included in the comparison allows the optimization of performance characteristics through null space movements that do not influence the task itself. The use of task redundancy to increase the performance characteristics compared to non-redundant kinematics must already take place in the dimensional synthesis of the structures, since the potential of this type of redundancy depends strongly on the dimensioning of the robot.The redundancy provides a degree of freedom for the solution of the inverse kinematics problem which influences the higher-level objective function of the dimensional synthesis. Since in the redundancy resolution one (possibly other) objective function is optimized, interactions between the cascaded optimizations (dimensional synthesis and redundancy resolution) have to be found to enable the two levels to complement each other. The focus of the proposed project is a transfer of methods from kinematic to task redundancy and the development of the optimization structures.