首页 /研究 /Recursive experimental structure re-design of a robot arm using rapid prototyping
OTHER

Recursive experimental structure re-design of a robot arm using rapid prototyping

Anton Pil, H. Harry Asada

发表年份
2002
引用次数
3

摘要

This paper describes an experimental recursive method for changing the hardware construction of a plastic robot arm in order to improve the system's dynamic performance. A system is implemented which allows the robot to improve its task performance through repetitive reinforcement of its structure. The structure reinforcement occurs using a coating method developed from stereolithographic rapid prototyping technology. A recursive algorithm, based on a gradient descent method and a parameter estimation theory, is developed to determine the optimal incremental structure changes. The resulting structure changes are then immediately executed by the stereolithographic reinforcement coating method. The entire process consisting of task execution, data analysis, and structure reinforcement is repeated until the task performance is improved to the desired level. An experimental system implementing the proposed technique was developed for building a high speed robot. The system is applied successfully in building a polycarbonate robot arm, showing a significant improvement in the arm dynamics.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>

关键词

Reinforcement learningComputer scienceTask (project management)RobotRobotic armRapid prototypingProcess (computing)ReinforcementRoboticsArtificial intelligence

相关论文

查看 OTHER 分类全部论文