Bayes分析中多源验前信息融合的ML-II方法

在Bayes小子样理论中,验前分布的获取和表示是一个关键问题.针对工程实践中遇到的验前信息的多源性,给出了一种基于第二类极大似然估计原理(M L-II)的多源异总体验前分布的融合方法,并通过仿真实例证明了该方法的有效性.     

基于卡尔曼滤波的信息融合算法优化研究

通过比较采用联邦卡尔曼滤波的状态向量融合和量测信息融合,得出量测信息融合优于状态向量融合,因为只有当卡尔曼滤波一致时状态向量融合才有效.采用基于最小均方差估计的观测值加权融合法融合了多传感器数据,保持了观测向量的维数.这种方法具有高效性.为了提高该算法的速度和精度,对系统的量测空间进行了等价变换,而等价系统的状态空间却没有改变.给出了等价变换前后的系统误差方差阵和状态估计均一致性的证明.把矩阵分析中的L-D分解算法运用到该算法中以避免计算矩阵的逆,从而改善了算法的稳定性和精度.举例验证了所设计算法的这些优点,给出了采用联邦卡尔曼滤波和所优化滤波算法的状态估计和误差的仿真结果,并依次进行了分析.经过这种优化,算法的精度和速度得到很大提高,已经应用到实际工程中.     

Tritium Burn-up Depth and Tritium Break-Even Time

Similarly to but quite different from the xenon poisoning effects resulting from fission-produced iodine during the restart-up process of a fission reactor, we introduce a completely new concept of the tritium burn-up depth and tritium break-even time in the fusion energy research area. To show what the least required amount of tritium storage is used to start up a fusion reactor and how long a time the fusion reactor needs to be operated for achieving the tritium break-even during the initial start-up phase due to the finite tritium breeding time that is dependent on the tritium breeder, specific structure of breeding zone, layout of coolant flow pipe, tritium recovery scheme, extraction process, the tritium retention of reactor components, unrecoverable tritium fraction in breeder, leakage to the inertial gas container, and the natural decay etc., we describe this new phenomenon and answer this problem by setting up and by solving a set of equations, which express a dynamic subsystem model of the tritium inventory evolution in a fusion experimental breeder （FEB）. It is found that the tritium burn-up depth is 317g and the tritium break-even time is approximately 240 full power days for FEB designed detail configuration and it is also found that after one-year operation, the tritium storage reaches 1.18kg that is more than the least required amount of tritium storage to start up three of FEB-like fusion reactors.     

基于区域特性量测的图像融合方法

在图像融合过程中，融合规则及融合算子的选择对于融合的质量至关重要，也是图像融合中至今尚未很好解决的难点问题。本文给出了一种新的融合规则——基于区域特性量测的融合规则，即在对某一分解层图像进行融合处理时，为了确定融合后的像素不仅要考虑参加融合图像中对应的各像素，而且要考虑参加融合像素的局部领域。     

彩色空间变换法在图像融合中的应用

本文讨论了ＲＧＢ空间与ＨＩＳ空间的特点及彩色空间变换的基本原理，并以卫星遥感获得的多光谱图像为例，成功的将彩色空间变换应用于空间信息与光谱信息的融合中，使融合后的图像同时具有较为丰富的空间信息和光谱信息，并对ＨＩＳ空间中Ｈ，Ｓ分量的灵活应用作了进一步的研究试验     

多量级多向梯度海空复杂背景红外弱点目标检测

由于红外图像一般带有较大的噪声，采用传统的目标检测方法效果不理想。本文提出了一种新的海空背景下受强杂波和噪声污染的红外图像弱点目标检测算法。用多个量级梯度对图像目标进行检测，并对检测的结果进行了表决融合。结果表明，基于表决融合的多量级多向梯度检测消除了云层、海浪和海天线等背景干扰，在实现高检测概率的同时，不仅可以达到较低的虚警概率，而且可检测信杂比为1的点目标。     

基于目标识别的多传感器图像信息融合技术

本文针对目标识别问题，论述了多传感器目标属性融合技术。包括目标属性融合概念、原理、处理结构、算法及应用举例。指出了多传感器目标属性融合中存在的主要问题及解决方法，并预测了发展趋势     

Preliminary Design of ITER Test Blanket Module HC-SB Structure

The helium cooled solid breeder blanket that represents the main stream in relevant blanket concepts. It is considered as one of the main options for the fusion reactor. This concept have advantage as follows： （ 1 ） structure simpleness and stability, （ 2 ） no effect of MHD, （ 3 ） a good compatibility in different materials.     

Fusion Reaction of ^16O＋^14N and Its Implication for the Production of ^26Al in Explosive Oxygen Burning

We suggest that the fusion reaction ^16O＋^14N may be a new way to produce ^26Al in interstellar medium. Adopting different mixing modes, we investigate the impact on the production of ^26Al in explosive oxygen burning and find that the result is extremely sensitive to mixing mechanisms. In some cases, we obtain an encouraging result, for example, the greatest final abundance of ^26Al reaches 7.779×10^-6, which means that the explosive oxygen burning may be a new origin of ^26 Al.     

PEG诱导的柱胞鱼腥藻（Anabaenacylindrica）原生质球融合

以多盐复合液配制的ＰＥＧ溶液可诱导柱胞鱼腥藻原生质球融合，融合率约５％，融合时间６～４０ｍｉｎ，可发生两个或多个原生质球的融合，ＰＥＧ使用浓度１０％