基于SIFT,K-Means和LDA的图像检索算法

图像检索一直是信息检索领域的难题。提出了一种基于尺度不变特征变换（SIFT,Scale Invariant Feature Transform）,K-Means和潜在狄利克雷分布（LDA,Latent Dirichlet Allocation）的图像检索算法。算法主要分为两个阶段。预备工作得到分类完成的图库、概率分配参数表和基本词库;实现检索是在预备工作的基础上归类测试图片,然后在该类下搜索最相似图片。对比传统的基于文本或内容的检索方法,该算法在检索之前将图片库中所有图片按其本身特征进行自动分类,取代人工标注图像信息的过程,同时由于整个算法完全基于图像特征,故此方法不会引入人工因素的干扰。实验结果表明,该算法能够较为准确地将要检索的图片归为图片库对应的类别中,有效地提高图像检索效率。     

基于WPI的多操纵面飞机积分滑模容错控制

针对含位置和速率限制、故障重构存在误差和时滞下的过驱动飞行器损伤故障的容错控制问题,提出了一种基于动态自适应加权伪逆法(WPI,Weighted Pseudo Inverse)的积分滑模主动容错方法.采用指令限制模块对饱和控制信号及瞬时干扰进行限制,设计了动态自适应控制分配律逐步减小指令饱和,同时通过故障估计值修正控制分配律来直接补偿气动力损失,降低了故障对系统稳定的影响;设计积分滑模律实现了含重构不匹配和时滞的稳定控制.仿真结果表明,所设计的控制器能快速准确地跟踪参考指令,对时滞具有较强的鲁棒性,同时对损伤故障具有较强容错能力.     

空战中机群编队分层优化算法

针对机群编队优化计算复杂的问题,提出了一种分层优化算法。根据敌我双方的距离、角度、速度以及飞机导弹、雷达的性能,建立了多机协同任务分配模型。根据空战中常用的基本飞机队形,对敌方机群编队进行分层,对每层分别计算己方对选取各种基本队形时的任务分配结果和队形优化优势值,通过比较得到己方机群编队的每层最优队形,当得到己方每层的最优队形之后,将己方每层最优队形进行组合解码就可得到己方机群编队的最优队形。仿真结果表明该方法能有效地解决机群编队队形优化问题,并且该算法具有较好的实时性。     

多操纵面飞机综合重构飞行控制方法

为提高飞机飞行安全性能,将被动重构控制与主动重构控制相结合,提出了一种综合重构飞行控制系统设计方法.采用轨迹线性化控制方法进行飞行控制律设计,使系统具有较强的鲁棒性.针对多操纵面飞机,提出面向故障的操纵面管理概念,通过基排序最优控制分配方法以基底的形式对操纵面进行重新组合,并按照期望控制目标进行排序,实现在操纵面故障下的控制重构.仿真结果表明,该控制系统能有效处理各种典型操纵面故障,在操纵面发生故障时仍能快速跟踪控制指令,保证较好的飞行性能.     

资源分配的柔性决策方法

资源分配问题是具有一个上级决策者和多个下级决策者的多目标柔性决策问题。为解决这样的问题,本文根据资源分配问题的特点,建立了柔性决策模型,并提出逐点线性化算法,以求解资源分配方案;提出用“承包机制”的激励策略作为引导下级决策者汇报真实信息的策略,这种策略不仅对真实信息是ε-可诱导的,而且也是鼓励下级努力工作的策略。     

大迎角综合飞行/推进控制系统设计与仿真

以先进推力矢量战斗机为背景,构建了基于力矩解算和分配的大迎角综合飞行/推进控制系统结构方案.在此基础上采用模糊方法设计推进稳定性控制系统;以逆动力学方法计算期望力矩;以饱和控制分配原则协调气动舵面和推力矢量的控制.仿真结果表明大迎角下推进系统可稳定工作,综合飞/推控制系统能合理协调气动舵面和推力矢量,控制系统能够快速准确的跟踪指令,完成机动动作.      

多目标非线性控制分配方法研究

 提出了一种新的多目标非线性规划控制分配方法,对其理论基础、建模与求解过程进行了详尽描述, 并给出了多目标非线性控制分配的解的评价指标及评价方法。该方法不依赖于伪控制指令与广义操纵面偏转角度之间呈线性关系这一假设,直接处理非线性控制分配问题,并充分利用冗余操纵面,实现不同飞行条件和任务下对多种目标的综合权衡分配。仿真结果表明该方法具有较高的精度,能很好地处理故障情况下的控制分配问题,为控制重构提供了算法保证。     

A hazard analysis via an improved timed colored petri net with time–space coupling safety constraint

Petri nets are graphical and mathematical tools that are applicable to many systems for modeling, simulation, and analysis. With the emergence of the concept of partitioning in time and space domains proposed in avionics application standard software interface(ARINC 653), it has become difficult to analyze time–space coupling hazards resulting from resource partitioning using classical or advanced Petri nets. In this paper, we propose a time–space coupling safety constraint and an improved timed colored Petri net with imposed time–space coupling safety constraints(TCCP-NET) to fill this requirement gap. Time–space coupling hazard analysis is conducted in three steps: specification modeling, simulation execution, and results analysis. A TCCP-NET is employed to model and analyze integrated modular avionics(IMA), a real-time, safety-critical system. The analysis results are used to verify whether there exist time–space coupling hazards at runtime. The method we propose demonstrates superior modeling of safety-critical real-time systems as it can specify resource allocations in both time and space domains. TCCP-NETs can effectively detect underlying time–space coupling hazards.     

常用文件系统的比较及转换

介绍了不同类型的文件系统以及它们与多种操作系统之间的兼容性，提供了实现各类文件系统之间相互转换的有效方法。     

Reducing mission operations costs through spacecraft autonomy: the near earth asteroid rendezvous (near) experience

With a maximum time of 12 days out of ground contact and a round-trip light time as high as 56 minutes, The Near Earth Asteroid Rendezvous (NEAR) spacecraft requires a moderate degree of onboard autonomy to react to faults and safe the spacecraft. Beyond the basic safing requirements, additional functions are carried out onboard. For example, on-board calculation of the Sun, Earth, asteroid, and spacecraft positions allow the spacecraft to autonomously orient itself for science and downlink operations. On-board autonomous momentum management during cruise relieves Mission Operations from planning, scheduling, and carrying out many manual momentum dumps. During development, additional operations, such as center-of-mass management during propulsive maneuvers and optical navigation were also considered for onboard autonomy on the NEAR spacecraft, but were not selected. The allocation of functions to onboard software or to ground operations involved tradeoffs such as development time for onboard software versus ground software, resource management, life cycle costs, and spacecraft safety.After two years of cruise operations, considerable experience with the NEAR autonomy system has accrued. The utility of some autonomous capabilities is greater than expected, others less so. Software uploads increased spacecraft autonomy in some cases, and the impact on Mission Operations can be assessed. Allocation of functions between spacecraft autonomy and ground operation during development of future missions can be improved by applying the lessons learned from the NEAR experience.