A5014984172- Robot Manipulation and Learning
- Robotic Path Planning Algorithms
- Soft Robotics and Applications
- Reinforcement Learning in Robotics
- Robotic Mechanisms and Dynamics
- Space Satellite Systems and Control
- Spaceflight effects on biology
- Robotic Locomotion and Control
- Prosthetics and Rehabilitation Robotics
- Space Exploration and Technology
- Fuel Cells and Related Materials
- Iterative Learning Control Systems
- Modular Robots and Swarm Intelligence
- Teleoperation and Haptic Systems
- Assembly Line Balancing Optimization
- BIM and Construction Integration
- Mechatronics Education and Applications
Beijing Institute of Technology
2017-2023
State Key Laboratory of Vehicle NVH and Safety Technology
2023
Shibaura Institute of Technology
2017-2019
Beijing Advanced Sciences and Innovation Center
2018
Ministry of Education of the People's Republic of China
2018
Using robots to assist or even replace rescuers for searching and rescuing has always been a research hotspot. Robots that can carry out dexterous operations at the scene are of great significance reduce life threat rescue workers. Aiming characteristics narrow space frequent physical interaction in scene, dual-arm mobile robot BIT-DMR composed humanoid upper body tracked platform is developed this letter. The main work as follows: a) A high-performance lightweight heavy payload...
On-orbit assembly has become a crucial aspect of space operations, where the manipulator frequently and directly interacts with objects in complex process. The traditional control limitations adapting to diverse tasks is vulnerable vibration, leading failure. To address this issue, we propose human-like variable admittance method based on damping characteristics human arm. By collecting velocity contact force arm operations assembly, analyze change establish active compliance model S-type...
Using space stations for a large number of observation, exploration, and research is necessary way to fully develop technology. It means experiment install the extravehicular experimental load by using plate. However, environment full danger, which poses threat health even safety astronauts. robots replace astronauts complete this task can effectively reduce Aiming at problem that configurations existing have difficulty in balancing contradiction between complexity dexterity, our previous...
This paper proposes a novel chameleon-like astronaut robot that is designed to assist, or even substitute, human astronauts in space station complete dangerous and prolonged work, such as maintenance of solar panels, so on. The can move outside the freely via hundreds aluminum handrails, which are provided help move. weighs 30 kilograms, consists torso, three identical 4-degree freedom (DOF) arms, end effectors, monocular vision system on each effector. Via multi-arm associated motion,...
In this study, we proposes a humanoid dual-arm explosive ordnance disposal (EOD) robot design. First, seven-degree-of-freedom high-performance collaborative and flexible manipulator is developed, aiming at the transfer dexterous operation of dangerous objects in EOD tasks. Furthermore, an immersive operated (FC-EODR) designed, which has high passability to complex terrains such as low walls, slope roads, stairs. It can remotely detect, manipulate, remove explosives environments through...
The field of artificial intelligence (AI) has advanced significantly over the years. One its achievements is deep reinforcement learning algorithm using which AI can play some Atari 2600 games better than humans. In this paper, optimal route construction machines such as bulldozers modeled based on learning. aim study to apply a grading machine enable it grade various surface types autonomously. A simple simulator created simulate task. addition, overall scenario made visible network by...
Deep reinforcement learning algorithms are rapidly growing, and expected to be applied many industrial fields. In this paper, we proposed a method that combines deep Q-network with batch normalization generate an optimal route for grading machine. The goal is achieve autonomous operation of the For platform, simulator was developed emulate work. evaluated simulator, showed better searching performance results than conventional method.
In recent years, deep reinforcement learning has better control performance than human in some Atari 2600 games. But it rarely any practical examples with this method the real world. paper, we propose to use construction domain. We cooperate JAXA and are a project make base moon. It is hard send worker operate leveling machine. So, our aim using machine can level ground autonomously various situations. simulate action simulator evaluate method. Also, comes prohibitive computational cost...
Traditional satellite assembly process mainly relies on manpower, so that there is an urgent need to introduce the robot system which able assist workers assembling satellites. The draws application of technology and specifications industrial robots other manipulators in relevant industries. In this paper, impedance controller used control movement response human's traction. influence parameters operation analyzed by simulation. Results experiment shows has good compliance external force...
After it was reported that an AI player scored higher in Atari2600 games than skilled human players by using deep reinforcement learning techniques, many researchers were inspired to apply leaning various fields. This paper focuses on the autonomous ground leveling work a bulldozer, which is expected optimize action of bulldozer. In previous work, we implemented Q method giving images as input data for network. However, when image convolution layer learning, requires large computational cost...