基于改进的概率假设密度多目标跟踪算法

针对多个目标相互紧邻时,概率假设密度滤波器难以正确估计当前目标个数以及目标状态问题,提出一种改进的高斯混合概率假设密度滤波算法。根据每一时刻更新后所有目标的权值构造权值矩阵,通过权值矩阵中目标权值的分布来检测当前目标权值是否存在更新错误。基于新的目标权值再分配策略,对权值矩阵中每个目标可能不正确的权值进行调整,使得每个目标能够获得合理的权值。仿真实验表明,该算法能够准确地估计紧邻目标数目以及状态。     

粒子滤波匹配追踪重构算法

针对现有贪婪迭代类压缩感知重构算法对非高斯量测噪声抵抗性差的问题,提出一种盲稀疏度下粒子滤波匹配追踪稀疏信号重构算法。该算法将鲁棒性更高的Huber损失函数替代常规的二次损失函数,用来增加对非高斯噪声的抵抗能力;并引入粒子滤波实现对原始信号的最优估计,以削弱量测噪声的影响;在信号稀疏度未知的条件下,结合稀疏度自适应匹配追踪算法实现盲稀疏度下的原信号重构。理论分析和仿真结果表明,所提算法可以有效抵抗因非高斯噪声干扰或稀疏度未知导致的重构精度降低,且重构性能优于现有典型贪婪迭代类算法。     

不同转位方案下双轴旋转惯导系统精对准分析

针对双轴旋转惯导系统转位方案的设计和选择问题,提出了3种合理的转位方案:16次序、24次序和32次序,建立了旋转惯导系统的误差方程和卡尔曼滤波模型.观测量取速度误差和位置误差时,分别对3种转位方案下精对准过程进行了分析仿真.结果表明,24次序转位方案精对准过程所需时间最短,精度最高,其次是32次序,16次序转位方案精对准过程所需时间最长,精度最低.     

SPEED BUMP ON THE ROAD TO GLOBAL ZERO

The prospect of the United States continuing to reduce the size of its nuclear arsenal to “very low numbers” has raised questions in Japan and South Korea, where US extended deterrence guarantees are premised on the “nuclear umbrella.” In both countries, however, concerns focus less on numerical arsenal size than on the sufficiency of specific nuclear and non-nuclear capabilities to meet evolving threats and on the degree of broader US commitment to these alliances. This article assesses developments in US-Japan and US-South Korea relationships in response to the Obama administration's nuclear disarmament policies, focusing on how the evolutionary course of those relationships may in turn condition prospects for sustaining this US nuclear policy direction. The analysis finds that the challenges of deterrence credibility and allied reassurance are difficult and long-term, but also that US nuclear arsenal size is secondary to broader political, strategic, and military factors in meeting these challenges. The evaluation concludes that strong alliance relationships and strategic stability in East Asia can be maintained while the size of the US nuclear arsenal continues to decline, but also that deterioration of these relationships could imperil core US nuclear policy and nonproliferation objectives.     

Formulations for dynamic lot sizing with service levels

In this article, we study deterministic dynamic lot‐sizing problems with a service‐level constraint on the total number of periods in which backlogs can occur over a finite planning horizon. We give a natural mixed integer programming formulation for the single item problem (LS‐SL‐I) and study the structure of its solution. We show that an optimal solution to this problem can be found in \begin{align*}\mathcal O(n^2\kappa)\end{align*} time, where n is the planning horizon and \begin{align*}\kappa=\mathcal O(n)\end{align*} is the maximum number of periods in which demand can be backlogged. Using the proposed shortest path algorithms, we develop alternative tight extended formulations for LS‐SL‐I and one of its relaxations, which we refer to as uncapacitated lot sizing with setups for stocks and backlogs. {We show that this relaxation also appears as a substructure in a lot‐sizing problem which limits the total amount of a period's demand met from a later period, across all periods.} We report computational results that compare the natural and extended formulations on multi‐item service‐level constrained instances. © 2013 Wiley Periodicals, Inc. Naval Research Logistics, 2013     

基于磁强计和光纤陀螺的小卫星姿态确定非线性滤波算法

提出了一种小卫星姿态确定的非线性滤波算法,该算法利用三轴磁强计和光纤陀螺作为姿态敏感器。在非线性滤波器的设计中,从两个方面对平方根sigma点卡尔曼滤波方法进行改进。第一,把姿态四元数的矢量部分、光纤陀螺的漂移和噪声组合,得到滤波器的增广状态向量;第二,分别建立向量旋转模型、最优化模型和误差四元数乘法模型来确保非线性滤波过程中四元数的归一化约束。仿真分析结果表明,本文提出的非线性滤波算法能够有效地提高小卫星的定姿性能,与扩展卡尔曼滤波相比,具有较高的精度、稳定性和较快的收敛速度;与无迹卡尔曼滤波相比,收敛性相当,但是精度略优,稳定性和计算效率较高。     

距离伪量测的空基平台单站无源定位算法

针对空中观测单站平台对海面、地面目标的测向定位问题,提出一种基于距离伪量测的最小二乘-不敏卡尔曼滤波(LSE-UKF)两阶段滤波算法。首先利用单次的测角信息以及目标离地高度的约束条件计算出目标的距离,采用最小二乘法对距离量进行滤波,得到较为准确的距离量。进一步将滤波所得距离量视为伪量测信息,结合观测所得的角度信息进行不敏卡尔曼滤波。仿真实验表明该算法提高了定位精度并且具有较快收敛速度。     

基于观测预测粒子滤波的雷达组网目标状态估计

雷达组网系统跟踪目标时,观测数据与目标跟踪状态成严重非线性关系,难以用卡尔曼滤波最优估计方法,处理非高斯非线性系统滤波估计问题的粒子滤波算法容易产生粒子退化问题。因此,使用观测预测粒子滤波算法解决这个问题,该算法基于观测似然进行重要性采样,结合一步预测信息计算粒子权值,保证了采样粒子处于高观测似然区,并充分利用了一步预测信息。仿真验证表明,将观测预测粒子滤波算法应用于目标状态估计,避免了粒子退化,收敛快,估计精度高。     

基于UKF的自适应网格交互式多模型算法

针对非线性观测条件下的机动目标跟踪问题,基于机动目标的协同转弯（CT）模型,采用无迹卡尔曼滤波（UKF）方法和自适应网格（AG）的模型集自适应策略,研究了一种变结构交互式多模型算法,即：基于UKF的自适应网格交互式多模型（UKF-AGIMM）算法。对二维机动目标跟踪的仿真结果表明,本算法与相应的固定结构交互式多模型算法相比,可以解决固定结构多模型算法存在的问题,有效提高多模型算法的精度和费效比,缩短计算时间,且适合工程应用     

结合非负张量表示与扩展隐Dirichlet分配模型的图像标注

由于"语义鸿沟"的存在,自动图像标注是一项极具挑战性的工作。考虑到图像低层视觉特征与高层语义概念的差异,分别从图像表示与语义建模两个方面来实现自动图像标注。在图像表示方面,提出了一种正则化约束下的非负张量表示方法,用以提取符合人眼视觉直观理解的图像高阶结构特征。在语义建模方面,提出了一种三层贝叶斯模型——扩展隐Dirichlet分配。该模型利用隐变量来实现图像与标注词的关联,并通过一种基于变分推理的期望最大值方法来估计参数。实验结果表明,ELDA模型在大规模数据库NUS-WIDE上的标注结果相较于现有方法有了显著的提高。