基于AUV垂直面运动控制的状态增减多模型切换

自主水下机器人(autonomous underwater vehicle,AUV)系统垂直面运动具有强耦合性、非线性等特点导致AUV系统垂直面难以精确控制,且其纵倾角需控制在某一阈值内,以保证系统安全。根据以上控制要求与AUV垂直面运动模型特点,提出将其解耦为深度控制子模型与纵倾角控制子模型,构建AUV系统垂直面运动控制模型集,根据模型集所含两子模型特点,提出了状态变量增减的多模型切换的概念。基于切换系统稳定判据——共同Lyapunov函数法,提出了变加权多模型切换策略,实现了切换过程的稳定性,避免了切换过程的抖动问题。多次湖泊试验证明,加权多模型切换策略避免了由于期望深度变化而造成的执行机构饱和、深度和纵倾角超调量过大等问题,且改善了AUV系统深度运动的静动态控制品质。

Beginning of a New Revolution of Science ＆ Technology

Modern science and technology is being developed towards destroying human and human calls for a new era of science ＆ technology. Contemporary science is seeing a significant breakthrough and a new revolution of science and technology is beginning. The theoretical basis of the new revolution of science and technology is ＂state space-time＂, ＂the unity of physics and biology＂, ＂thinking motion and automatic organization of the universe＂ and ＂thinking＇s active control＂. The new revolution of science and technology takes material （microscopic particle） science, life （organic and inorganic life） science and thinking science as the fundamental sciences, takes life technology, new information technology, new material technology and new energy technology as the basis of the overall technological system, progresses in all aspects, including life technology, medicine, new material technology, new information technology, new energy technology, new robot technology, new manufacturing technology, laser technology, astronavigation technology, ocean technology, military technology, new educational technology, etc., and at the same time Qigong technology with thinking control as its core will emerge. In the era of life, thinking and Qigong technology, technology proceeds very rapidly, global society and economy develop at a high speed, and globalization is accelerated. Life, thinking and Qigong technologies feature improvement of human internal energy and quality, purify human soul, reduce material desire, and greatly reduce environmental disruption. The overall （thinking and Qigong） technology takes the establishment of the cosmic era of human as its ultimate goal.

ULEO:表示合成实验规程的实验操作统一语言

【目的】面对智能科研与科学机器人对高质量实验规程数据的需求,解决合成实验规程中的实验操作词统一表示问题。【方法】综合利用数据和专家知识协同驱动的方式,从合成相关的论文与专利文本中识别并标准化实验操作词。实验操作词识别主要选用较为先进的开源大模型ChatGLM2-6B,实验操作词标准化则混合应用Wu-Palmer和余弦相似度,辅以专家经验知识判别分类的准确性。【结果】分别获取149个无机合成实验操作词和141个有机合成实验操作词,两者交集124个词。经判定在两类合成实验中分别出现的操作词中多数并不具备鲜明的类别特色,因此可取两类合成实验操作词的并集,共计166个,用于统一表示有机、无机及其杂化合成实验操作。【局限】仅使用基础的提示工程来激发大模型识别实验操作词,准确率有待提升;所用的数据主要源于当前免费公开的数据集,不够全面、丰富;仅关注合成、工程和基础步骤中涉及的操作词,未涉及动态、分析与命名反应中的操作词。【结论】本文构建一套表示合成实验操作的统一语言,用于表示有机、无机及其杂化合成反应中的实验操作,不同类型的合成实验操作词在表示上差异不大,在使用频次和倾向上确有不同,今后可据此优先选择研制科学机器人相应的实验操作功能。

An interview with Dr.Raj Reddy on artificial intelligence

Dr.Raj Reddy is the Moza Bint Nasser University Professor of Computer Science and Robotics in the School of Computer Science,Carnegie Mellon University,Pittsburgh,Pennsylvania,USA.He was awarded the ACM Turing Award in 1994.