基于战例数据的制胜要素选择方法

针对战争制胜机理定量研究需求,提出了一种基于混合评估的制胜要素选择算法。针对战例数据特点,该算法选择两种过滤方法分别从不同方面对要素全集进行评估排序而后加权得到综合排序结果;将结果作为遗传算法的初始种群,而后以分类精度作为个体适应度函数。选择几种典型的分类器综合比较,筛选出规模较小、性能较好的要素子集。测试结果表明,该算法不仅能有效地减少要素子集规模,还可以进一步提高制胜机理分析模型的准确率和效率。     

Single‐commodity stochastic network design under demand and topological uncertainties with insufficient data

Stochastic network design is fundamental to transportation and logistic problems in practice, yet faces new modeling and computational challenges resulted from heterogeneous sources of uncertainties and their unknown distributions given limited data. In this article, we design arcs in a network to optimize the cost of single‐commodity flows under random demand and arc disruptions. We minimize the network design cost plus cost associated with network performance under uncertainty evaluated by two schemes. The first scheme restricts demand and arc capacities in budgeted uncertainty sets and minimizes the worst‐case cost of supply generation and network flows for any possible realizations. The second scheme generates a finite set of samples from statistical information (e.g., moments) of data and minimizes the expected cost of supplies and flows, for which we bound the worst‐case cost using budgeted uncertainty sets. We develop cutting‐plane algorithms for solving the mixed‐integer nonlinear programming reformulations of the problem under the two schemes. We compare the computational efficacy of different approaches and analyze the results by testing diverse instances of random and real‐world networks. © 2017 Wiley Periodicals, Inc. Naval Research Logistics 64: 154–173, 2017     

基于圆形滤波器HOG特征的航拍图像车辆快速检测算法

航拍图像中车辆一般近似为矩形结构,因此通过统计检测窗口中的梯度方向直方图并根据梯度主方向估计车辆朝向,然后将检测窗口旋转到相应方向进行分类器判别。车辆检测采用级联boosting分类器和梯度方向直方图特征(HOG,Histograms of Orient Gradient),针对旋转窗口中HOG特征的计算,设计了一种基于圆形滤波器的HOG(Circle Filter based HOG,CFHOG)特征,与传统基于积分直方图的HOG特征提取方法相比,显著提高了旋转窗口中HOG特征的计算效率,此外在计算每个像素的梯度时采用查找表代替梯度向量的求模和角度计算也减小了计算量。使用真实图像进行的实验表明本文车辆检测算法快速高效。     

基于稀疏贝叶斯学习的扩展目标雷达关联成像*

雷达关联成像不依赖于雷达与目标的相对运动,是一种高分辨凝视成像方式。传统的关联成像方法未考虑复杂扩展目标的结构信息,在高分辨成像时的应用受到限制,为此提出一种自适应结构配对稀疏贝叶斯学习方法。该算法在稀疏贝叶斯学习的框架内针对扩展目标建立一种结构配对层次化高斯先验模型,然后采用变分贝叶斯期望-最大化算法交替进行目标重构和参数优化。该算法将某一信号分量的重构与周围信号分量联系起来,并能在迭代过程中自适应地调整表征各信号分量相关性的参数。实验结果表明,该方法针对扩展目标可以有效地进行高分辨成像。     

Missile defense and strategic stability: Terminal High Altitude Area Defense (THAAD) in South Korea

South Korea is threatened by its troubled relationship with North Korea. North Korea possesses a large cache of missiles as well as chemical and biological weapons, and the future potential to mount nuclear weapons on its missiles. The United States is also challenged because of its defense commitments to Seoul. As a countermeasure, the United States and South Korea decided to deploy Terminal High Altitude Area Defense (THAAD) missile defenses in South Korea. However, China has objected. Chinese scholars believe the THAAD radar would be able to track Chinese inter-continental ballistic missiles, thereby weakening their deterrent. A technical analysis does not support this assertion. However, it is vital for South Korea, given its proximity and economic interdependence, to reassure China. South Korea should highlight that THAAD will be deployed by the United States Forces Korea and is not a commitment by Seoul to become part of U.S.-led missile defenses in the Asia-Pacific.     

非椭圆扩展目标联合跟踪与分类算法

充分利用目标尺寸和形状信息,提出了一种基于星凸随机超曲面模型(random hypersurface model, RHM)的非椭圆扩展目标联合跟踪与分类(joint tracking and classification, JTC)算法。将目标空间扩展状态建模为星凸形状,通过目标类别相关先验信息的矢量化建模,建立起其与目标瞬时扩展状态的关系,并在统一的贝叶斯滤波框架下,实现跟踪与分类的一体化处理;进一步对目标运动学状态和扩展状态单独进行建模,并通过构建扩展状态的似然函数,利用粒子滤波实现目标类别概率算式的递推处理。仿真结果表明:与基于椭圆形状的扩展目标JTC算法相比,所提算法能对尺寸相近、形状不同的目标进行准确分类,同时可改善目标状态的估计效果;与基于星凸RHM的扩展目标跟踪算法相比,所提算法能大幅提高目标状态的估计性能。     

基于改进IMM-UKF的弹道导弹再入段跟踪数据滤波?

针对弹道系数未知的弹道导弹再入段跟踪雷达测量数据滤波这类非线性强的滤波问题,提出可变多模型无迹卡尔曼滤波算法。利用无迹卡尔曼滤波逼近精度高,计算量小,适应于任意非线性模型的特点,将其作为多模型的基本滤波器;滤波算法根据各模型正确描述目标状态的概率,动态地改变多模型数量和模型参数。上述方法的综合运用,提高对目标状态估计精度,降低了计算的复杂度,仿真实验验证了方法的有效性。     

基于概率图模型的海上多目标跟踪

针对海上复杂环境中多运动目标跟踪问题,提出了一种融合概率图模型和粒子滤波的跟踪方法.该方法采用二元马尔可夫随机场构建多目标模型,利用势函数表示其联合概率分布,最后用变差分法和粒子滤波推理后验概率密度.通过二阶回归模型准确构建状态转移方程,以及Mean-Shift迭代加速粒子采样,提高了算法精确度和运行效率.实验表明:该方法能很好地实现海上复杂环境中的多目标实时跟踪.     

用于消除缓变伪距偏差的抗差扩展卡尔曼滤波方法

针对卫星导航系统中伪距观测量存在缓变伪距偏差,导致定位精度下降的问题,提出了一种基于滑动积累的抗差扩展卡尔曼滤波方法。该方法在抗差扩展卡尔曼滤波算法的基础上,通过滑动窗口内多个历元残差矢量的累积,增大标准化残差矢量的伪距偏差,来提高对缓变伪距偏差故障的抗差能力。仿真表明,该方法能有效消除或者减弱微小伪距偏差和缓慢增长伪距偏差造成的位置误差,与传统抗差扩展卡尔曼滤波算法比较,定位精度得到明显改善。     

火箭弹卷弧尾翼的轻量化研究

针对火箭弹卷弧尾翼在特定气动载荷下的轻量化问题,基于结构动力学理论,依托ABAQUS非线性有限元软件,建立了火箭弹卷弧尾翼的有限元模型。通过对比碳纤维复合材料(CFRP)和合金钢材料的卷弧尾翼的有限元分析,得出了碳纤维复合材料在满足强度和刚度的要求下,使卷弧尾翼减重达80%。该方法为火箭弹卷弧尾翼的轻量化设计提供了一种参考和借鉴。