报告题目: A Sample Efficient Model-based Deep Reinforcement Learning Algorithm with Experience Replay for Robot Manipulation
报告地点：zoom在线会议 会议ID：939 3075 5234 密码： nMsa4a
For robot manipulation, reinforcement learning has provided an effective end to end approach in controlling the complicated dynamic system. Model-free reinforcement learning methods ignore the model of system dynamics and are limited to simple behavior control. By contrast, model-based methods can quickly reach optimal trajectory planning by building a dynamic system model. However, it is not easy to build an accurate and efficient system model with high generalization ability, especially when facing complex dynamic system and various manipulation tasks. Furthermore, when the rewards provided by the environment are sparse, the agent will also lose effective guidance and fail to optimize the policy efficiently, which result in considerably decreased sample efficiency. In this paper, a model-based deep reinforcement learning algorithm, in which a deep neural network model is utilized to simulate the system dynamics, is proposed for robot manipulation. The proposed deep neural network model is robust enough to deal with complex control tasks and possesses the generalization ability. Moreover, a curiosity-based experience replay method is incorporated to solve the sparse reward problem and improve the sample efficiency in reinforcement learning. The agent who manipulates a robotic hand, will be encouraged to explore optimal trajectories according to the failure experience. Simulation experiment results show the effectiveness of proposed method. Various manipulation tasks are achieved successfully in such a complex dynamic system and the sample efficiency gets improved even in a sparse reward environment, as the learning time gets reduced greatly.
报告人简介： 张成 副教授 博导 日本茨城大学
Cheng ZHANG received his Ph.D. degree from Waseda University, Tokyo, Japan, in 2015. From 2008 to 2015, he was a research engineer at Sony Digital Network Applications, Japan and HGST Japan, Inc. (ex Hitachi Global Storage Technologies), where he researched and developed control algorithms for image stabilization module of Sony digital camera, and servo control algorithms for next generation high capacity HDD. From 2015 to 2020, he was an assistant professor of Graduate Program for Embodiment Informatics (Program for Leading Graduate School) at Graduate School of Fundamental Science and Engineering, Waseda University. He is currently an assistant professor of Department of Mechanical System Engineering, Ibaraki University, Ibaraki, Japan. His research interests include communication network engineering, machine control algorithm, IoT robot, reinforcement learning, and embedded software. He received the IEICE Young Researcher's Award in 2013. He is a member of IEICE, IEEE and ACM.