Findings from Northwestern Polytechnic University Update Knowledge of Robotics ( Episode-fuzzy-coach Method for Fast Robot Skill Learning)
西北理工大学最新机器人知识发现（快速机器人技能学习的事件模糊教练法）
扫码查看
点击上方二维码区域，可以放大扫码查看
摘要
由一名新闻记者兼机器人与机器学习的新闻编辑每日新闻-一项关于机器人的新研究现在开始了。根据NewsR X记者在中国西安的新闻报道，研究表明：“为了在真实世界中实现机器人技能学习，需要将强化学习算法应用于连续问题中，提高样本效率。混合智能被认为是解决这一问题的有效方法，因为它能够利用人类的知识和经验加快学习过程。”本研究经费来源于国家自然科学基金(NSFC)。新闻记者引用了西北理工大学的一篇研究文章,“因此,我们提出了一种模仿人类模糊逻辑的事件模糊教练(COrrective Advi Communications by Humans),将人类的知识和经验融入到学习过程中,并通过人类反馈和用户设计的模糊规则来提供知识和经验。”摘要:结合路径策略改进(PI2),实现混合智能,实现机器人快速技能学习,并用本文提出的投掷动作原语来表示投掷动作策略,仿真结果表明,该方法的学习效率分别提高了72%和42.86%。与纯PI2和PI2+COACH相比,实验验证的our方法比PI 2+COACH的效率提高46.67%,实验结果表明,该方法的性能不受用户相关领域知识水平的影响。
Abstract
By a News Reporter-Staff News Editor at Robotics & Machine Learning Daily News Daily News – A new study on Robotics is now availab le. According to news reporting from Xi’an, People’s Republic of China, by NewsR x journalists, research stated, “To realize robot skill learning in the real wor ld, reinforcement learning algorithms need to be applied in continuous problems with high sample efficiency. Hybrid intelligence is regarded as an available sol ution for this problem, due to the ability to speed up the learning process with human knowledge and experience.” Financial support for this research came from National Natural Science Foundatio n of China (NSFC). The news correspondents obtained a quote from the research from Northwestern Pol ytechnic University, “Therefore, we propose Episode-Fuzzy-COACH (COrrective Advi ce Communicated by Humans), to imitate human fuzzy logic and involve human intel ligence in the learning process. In this framework, human knowledge and experien ce are involved in the learning process, which are provided by human feedback an d fuzzy rules designed by human users. Moreover, it is combined with Path Integr als Policy Improvement ( PI2 ), to realize hybrid intelligence, which is used to realize fast robot skill learning. Throwing Movement Primitives proposed in thi s article is used to represent the policy of ball-throwing skill. According to t he simulation results, the learning efficiency of our method is increased by 72% and 42.86%, respectively, compared with pure PI2 and PI2+ COACH. Ou r method validated in experiments is 46.67% more effective than PI 2+ COACH. The results also show that the performance of our method is not affect ed by users’ knowledge level of the related field.”
Key words
Xi’an/People’s Republic of China/Asia/Emerging Technologies/Machine Learning/Robot/Robotics/Northwestern Polytech nic University
NETL
NSTL