Reinforcement Learning Based Precise Positioning Method for a Millimeters-Sized Omnidirectional Mobile Microrobot

Authors:
Jianghao Li;Zhenbo Li;Jiapin Chen
Affiliations:
Key laboratory for Thin Film and Microfabrication of Ministry of Education, Institute of Micro and Nano Science and Technology, Shanghai Jiaotong University, Shanghai, China 200240 and College of ...;Key laboratory for Thin Film and Microfabrication of Ministry of Education, Institute of Micro and Nano Science and Technology, Shanghai Jiaotong University, Shanghai, China 200240;Key laboratory for Thin Film and Microfabrication of Ministry of Education, Institute of Micro and Nano Science and Technology, Shanghai Jiaotong University, Shanghai, China 200240
Venue:
ICIRA '08 Proceedings of the First International Conference on Intelligent Robotics and Applications: Part I
Year:
2008

Citing 2
Cited 1

Introduction to Reinforcement Learning

Introduction to Reinforcement Learning
Fundamental Principles and Issues of High-speed Piezoactuated Three-legged Motion for Miniature Robots Designed for Nanometer-scale Operations

International Journal of Robotics Research

Microassembly path planning using reinforcement learning for improving positioning accuracy of a 1 cm3 omni-directional mobile microrobot

Applied Intelligence

Quantified Score

Hi-index	0.00

Visualization

Abstract

A tri-wheeled omnidirectional microrobot actuated by 3mmelectromagnetic micromotors is described. Since it is designed formicroassembly in a microfactory, the positioning precision becomesa key parameter. However, with traditional PID control, themicrorobot positioning precision is not high during a bearing/axisassembly task. This paper presents a reinforcement learning (RL)algorithm based on the on-policy (Sarsa(λ )) methodusing linear function approximation. The algorithm is used togenerate an optimal path by controlling the choice of four movingactions of the microrobot. The aim is to reach the target positionwith high positioning precision. Simulations show that this RLalgorithm is able to greatly improve the positioning precision withregard to the global path optimization.