液体火箭发动机高速涡轮泵的振动故障检测

讨论了涡轮泵故障的几个主要原因，据此提取涡轮泵振动数据的特征，用BP神经网络的方法进行故障检测。BP神经网络的训练样本集由一个具有无监督聚类功能的神经网络从原始的特征向量集获取。     

二次奖罚学习自动机

研究了奖罚型学习自动机的一种非线性强化算法。与线性的奖罚模型（ＬＲＰ）不同,新模型的行动选择概率的更新函数为二次的。这使得该模型的学习性能优于ＬＲＰ,且对不同的环境,其具有不同的行为和特点。     

防空导弹武器系统故障诊断技术探究

对主要的故障诊断技术进行了系统地归纳和分类,并重点讨论了基于知识处理的智能故障诊断方法,对专家系统故障诊断方法、故障树故障诊断方法、模糊故障诊断方法、神经网络故障诊断方法以及基于支持向量机的故障诊断方法进行了详细介绍。     

Artificial equilibrium points for a solar balloon in the α Centauri system

This paper discusses the generation, stability, and control of artificial equilibrium points for a solar balloon spacecraft in the α Centauri A and B binary star system. The continuous propulsive acceleration provided by a solar balloon is shown to be able to modify the position of the (classical) Lagrangian equilibrium points of the three-body system on a locus whose geometrical form is known analytically. A linear stability analysis reveals that the new generated equilibrium points are usually unstable, but part of them can be stabilized with a simple feedback control logic.     

大型捆绑运载火箭姿态控制系统

本文介绍了大型捆绑运载火箭姿态控制系统方案、工作原理和首发飞行试验结果,并将部分参数同美国航天飞机进行比较。     

人工拉格朗日点附近的被动稳定飞行

利用太阳帆能在三体问题中实现人工拉格朗日点,人工拉格朗日点克服了经典拉格朗日点位置固定的缺点,研究人工拉格朗日点的被动控制对深空探测有重要的意义。理论上人工拉格朗日点都不稳定,研究表明在被动控制下存在某些人工拉格朗日点的稳定特性与稳定平衡点非常接近,在工程上可以认为稳定。被动控制可以通过设计太阳帆来实现,本文给出了被动稳定太阳帆的设计,在该设计下考虑轨道和姿态的耦合动力学方程。基于该耦合方程研究了人工拉格朗日点的稳定性。仿真结果表明被动太阳帆使得人工拉格朗日点稳定。     

基于进化神经网络的组合导航系统滤波算法研究

组合导航系统中,卡尔曼滤波算法的估计特性依赖于精确的系统模型和准确的外观测数据.任何一个条件不满足将导致滤波精度下降甚至发散.为此,本文引入进化神经网络和自适应卡尔曼滤波技术.仿真结果证明,文中所提算法能够有效克服常规卡尔曼滤波的缺点,并保持较高的精度.     

中国运载火箭技术发展

中国航天运输系统建设起步于20世纪60年代,经过近50年的发展,取得了举世瞩目的成就,建设了布局合理、覆盖全面的空间运输系统体系,能够将不同有效载荷发射到低、中、高不同轨道。国际合作方面,在搭载发射、商业卫星发射服务和在轨交付3个方面也取得了一定成绩。对中国航天运输系统发展成就进行了总结,对航天运输系统未来发展特别是人工智能技术应用进行了展望。     

智能制造环境下的数控系统发展需求

"智能"是"中国制造2025"的核心要素,数控系统是制造装备实现智能化的关键要素。论文首先提出机床的信息互通和开放式标准数据接口是实现机床智能化的前提,举例说明了STEP-NC在通用数据接口方面的优越性,概述了适用于数控系统的通讯总线特征;在此基础上,介绍了工艺大数据对生产加工的重要性和其对新一代数控系统的要求,针对制造业数据平台的需求,对大数据信息挖掘技术做了相关介绍。依托于以上两点,总结了智能化数控系统可具备的功能,依托于工艺大数据平台,重点分析了智能工艺规划和加工参数自整定的实现方案,以及智能数控对高速高精加工所提供的便利。     

Toward understanding the active SETI debate: Insights from risk communication and perception

Insights from the robust field of risk communication and perception have to date been almost totally absent from the policy debate regarding the relative risks and merits of Active SETI or Messaging to Extraterrestrial Intelligence (METI). For many years, the practice (or proposed practice) of Active SETI has generated a vigorous and sometimes heated policy debate within the scientific community. There have also been some negative reactions in the media toward the activities of those engaged in Active SETI. Risk communication is a scientific approach to communication regarding situations involving potentially sensitive or controversial situations in which there may be high public concern and low public trust. The discipline has found wide acceptance and utility in fields such as public health, industrial regulation and environmental protection. Insights from the scientific field of risk communication (such as omission bias, loss aversion, the availability heuristic, probability neglect, and the general human preference for voluntary over involuntary risks) may help those who have participated in either side of the debate over Active SETI to better understand why the debate has taken on this posture. Principles of risk communication and risk perception may also help those engaged in Active SETI to communicate more effectively with other scientists, the public, with the media, and with policy makers regarding their activities and to better understand and respond to concerns expressed regarding the activity.