Evaluation of a regional real-time precise positioning system based on GPS/BeiDou observations in Australia

The performance of real-time (RT) precise positioning can be improved by utilizing observations from multiple Global Navigation Satellite Systems (GNSS) instead of one particular system. Since the end of 2012, BeiDou, independently established by China, began to provide operational services for users in the Asia-Pacific regions. In this study, a regional RT precise positioning system is developed to evaluate the performance of GPS/BeiDou observations in Australia in providing high precision positioning services for users. Fixing three hourly updated satellite orbits, RT correction messages are generated and broadcasted by processing RT observation/navigation data streams from the national network of GNSS Continuously Operating Reference Stations in Australia (AUSCORS) at the server side. At the user side, RT PPP is realized by processing RT data streams and the RT correction messages received. RT clock offsets, for which the accuracy reached 0.07 and 0.28?ns for GPS and BeiDou, respectively, can be determined. Based on these corrections, an accuracy of 12.2, 30.0 and 45.6?cm in the North, East and Up directions was achieved for the BeiDou-only solution after 30 min while the GPS-only solution reached 5.1, 15.3 and 15.5?cm for the same components at the same time. A further improvement of 43.7, 36.9 and 45.0 percent in the three directions, respectively, was achieved for the combined GPS/BeiDou solution. After the initialization process, the North, East and Up positioning accuracies were 5.2, 8.1 and 17.8?cm, respectively, for the BeiDou-only solution, while 1.5, 3.0, and 4.7?cm for the GPS-only solution. However, we only noticed a 20.9% improvement in the East direction was obtained for the GPS/BeiDou solution, while no improvements in the other directions were detected. It is expected that such improvements may become bigger with the increasing accuracy of the BeiDou-only solution.     

Series of radiation-hardened, high efficiency converters for high voltage bus

A review of common switching techniques for dc/dc power conversion is presented. The evolution of the forward converter is discussed, leading naturally to the introduction of the double forward converter. Technical details for a series of radiation-hardened converters are then provided. This series was designed for space power management and distribution (PMAD) applications. The successful development of the PMAD converter demonstrated the viability of high voltage bus.     

高燃速丁羟推进剂配方研究

通过多种途径的试验,对高燃速丁羟推进剂配方进行了研究.试验结果表明,采用超细防结块氧化剂和液固组合燃速催化剂能使推进剂的燃速达到70mm/s以上(在6.864Mpa压强下);采用组合工艺助剂可改善推进剂工艺性能.本推进剂燃烧稳定,压强指数和温度敏感系数较低,力学性能良好,为高燃速推进剂的研制奠定了良好的基础.     

一种基于FPGA的超高速FFT处理器设计

通过采用合并字节的预整序单元、全并行的基4运算单元、首级存储单元外置的移位延时整序结构和逐级 扩充字长的定点算法,用FPGA实现了一个超高速的基4全并行流水线FFT处理器;在66MHz时钟频率下,处理速度可 达到256K点／ms。     

振动对空间光通信系统误码率影响的分析

在空间光通信系统中，由于卫星平台的振动，使得光发射机发生振动，从而对通信系统的误码率产生影响，本文针对光射光束为高斯光束时，振动对误码率的影响问题进行研究。结果表明，在振动时不仅振幅影响着误码率，而且波长也对误码率有影响，但通信距离对误码率的影响不大，一般可以忽略。     

基于分层式强化学习的移动机器人导航控制

针对未知环境下的移动机器人导航问题,本文提出了一种基于分层式强化学习的混合式控制方法。利用栅格-拓扑相结合的环境表示及地图学习方法,通过分层式强化学习在不同控制层次的扩展设计移动机器人的反应式和慎思式导航控制,实现了全局导航和局部导航控制的协调优化。实验及测试结果证明,该控制方法能实现导航任务的全局优化,避免陷入局部极小,并对未知动态环境具有较强的适应性。     

压电驱动自耦合射流流动特性实验

为了研究压电驱动狭缝喷口自耦合射流的流动特性,采用PIV、热线风速仪测试手段,对自耦合射流激发器在不同激励因素下的流场和速度场分布进行了实验和分析.结果显示,自耦合射流在狭缝出口处产生了反向涡对,随着自耦合射流的发展,射流呈现出在喷口短轴方向急剧向两侧扩展、而在喷口长轴方向先收缩后缓慢扩展的流动特征;自耦合射流的速度分布在法线方向呈现先上升后下降的趋势,在z/b=10左右速度达到最大值;在射流展向上,短轴方向速度呈规律的对称分布和速度自模的特征,而长轴方向速度近喷口区域呈现马鞍状分布,随着法向距离增加这种趋势消失.研究中发现,激发器存在两个谐振频率,在谐振频率激发下自耦合射流的速度和涡量比较大.与常规射流相比,自耦合射流显示出了独特的流动特性.     

数字地形数据的二维三次卷积插值

由于数字地形数据库是以网格的形式给出网格点上的地形高度信息，在低空突防轨迹优化过程中，往往需要非网格蹼地形的高度信息和地形曲面的导数，这些都是离散的地形数据库所无法提供的。对二维的地形数据进行插值是解决问题的有效手段。本文提出一种二维离散数据的插值方法－－二维三次卷积插值。该插值函数是地形曲面的三次逼近，具有连续的三阶导数，可以非常好地满足低空突防轨迹优化的需要。仿真结果表明，它具有比双线性插值和     

提高图像轮廓清晰度的快速插值放大方法

分析了当前主要的高清晰度图像插值放大方法,为实现快速放大而选择了基于图像轮廓插值改善清晰度的技术路线,并在已有算法的基础上开展了进一步的研究.提出了新的图像轮廓线光顺数值算法,对轮廓线划分的两类子区域提出了精度更高的插值算法.最后,以典型平面图形的图像和由3DMax生成的三维渲染图像为例对本文方法进行检验,与双线性插值结果的对比说明,该方法可以明显改善插值放大效果.     

协同多目标攻击空战决策的启发式粒子群优化算法

利用协同多目标攻击战术的特定知识，并结合粒子群算法，提出了一种用于空战决策的启发式粒子群算法。该算法利用粒子群算法对解空间探索能力强，容易跳出局部最优陷井及启发式算法局部搜索能力强的优点，快速、高效地对全局最优值进行搜索。该算法通过求解友机导弹对目标的最优分配来确定空战决策方案。仿真实验结果表明。本文算法对最优空战决策方案的搜索性能明显优于普通粒子群算法及其他两种遗传算法。