参照化流形空间融合学习的敏感特征提取与异常检测方法

针对遥测振动信号冲击强、响应周期短、共振频带宽和小样本等特点导致异类模式识别率低的问题,提出基于参照化流形空间融合学习的敏感特征提取与异常检测方法。采用多尺度分析方法将信号正交无遗漏地分解到各尺度带中,提取多尺度特征构造高维特征集;以相同的正常信号样本结合相同类型的异常样本建立专属参照化模型单元,采用线性流形学习获取各参照化模型单元多尺度流形特征差异,增强异常特征的敏感性。融合各参照化模型单元的投影矩阵对原始特征集进行升维再学习,获取低维多尺度敏感流形特征;输入到分类器实现对未知样本状态辨识。实测信号处理结果验证了算法的有效性。     

Liquid phase in-situ synthesis of LiF coated boron powder composite and performance study

Among practical metal additives, boron (B) has a high volumetric heating value, making it a promising choice as a fuel additive. Although B can theoretically yield a large amount of energy upon complete combustion, its combustion is retarded by the initial presence of B oxide, which coats the surfaces of B particle. To improve the ignition and combustion properties of B powder, LiOH and NH₄F were used as precursors to synthesize uniformly LiF-coated B composites (LiF-B) in situ. The LiF-B mixture was also prepared for comparison using a physical method. X-ray diffraction (XRD), Fourier-transform infrared (FTIR), scanning electron microscope (SEM), and energy-dispersive X-ray spectroscopy (EDS) were used to characterize the morphologies and compositions of the products. The thermal and combustion properties of the samples were characterized by thermal gravity-differential thermal gravity (TG-DTG), differential scanning calorimetry (DSC) and closed bomb experiment. The XRD, FTIR, SEM and EDS results demonstrated the successful preparation of the coated LiF-B sample. The TG-DTG and closed bomb experiment results indicated that the addition of LiF decreased the ignition temperature of B powder, and increasing its reaction efficiency. DSC results show that when LiF-B was added, the released heat of underwater explosive increased by 6727.2, 7280.4 and 3109.6 J/g at heating rates of 5, 10, and 15 °C/min, respectively. Moreover, LiF-B decreased the activation energy of secondary combustion reaction of explosive system as calculated through Kissinger's method by 28.9%, which indicated an excellent catalytic effect for the thermal decomposition of underwater explosive. The results reveal that LiF can improve the combustion efficiency of B powder, thereby increasing the total energy of explosives. The mechanical sensitivity increased slightly after adding LiF-B to the underwater explosive. Compared to the underwater explosive with added B, the mechanical sensitivity of the explosive with added LiF-B was significantly lower.     

希尔伯特-黄变换在脉冲涡流信号消噪与识别中的应用

针对传统方法无法有效识别不同尺寸细小裂纹所产生的脉冲涡流信号,提出一种基于希尔伯特-黄变换的脉冲涡流信号消噪与识别算法。对脉冲涡流信号进行集成经验模态分解并通过归一化自相关函数及其方差特性分选出含有噪声的本征模态函数;对含噪声的本征模态函数进行阈值消噪并与未做处理的本征模态函数重构成无噪声信号;对无噪声信号进行希尔伯特-黄变换并计算出希尔伯特边际谱;根据希尔伯特边际谱的差异识别出不同细小尺寸的表面与下表面裂纹。实验结果表明了所提方法的有效性,经过集成经验模态分解消噪,消除了噪声对脉冲涡流信号的干扰;而基于希尔伯特-黄变换的方法则能够有效识别出不同尺寸的裂纹。     

作战使命分解与任务建模方法

为解决联合作战计划的多样性、复杂性与不确定性,在计划的基础工作上,定义了基本概念,包括作战想定、作战使命、子任务与元任务,提出了任务描述方法、使命分解原则与分解方法,实现了不同分解方法在结果上的同一性,为作战计划的标准化、程序化与智能化奠定了基础.     

信源数动态变化的一种稳健超定盲分离算法

针对大多数现有盲分离算法都假定信源数已知,分析了信源数未知且动态变化时的超定盲分离(接收天线数多于信源数)算法,通过奇异值分解在线估计信源个数,并修正自适应神经网络的维数。文章在经典Cichocki-Unbenauen算法中加入动量项,设置可变步长,使分离性能得到较大的改进,同时避免训练陷入局部最小。与文献[1]新提出的ANA算法相比,仿真结果证明本文算法具有更快的收敛速度和更小的稳态误差。     

飞机发动机关键系统智能故障诊断方法

飞机发动机是一个非常复杂的大系统,由于其结构复杂,工作环境恶劣,对其关键系统的故障进行准确诊断始终是困扰业界的技术瓶颈之一。提出了采用EMD小波阈值降噪与主元分析相结合的方法,对飞机发动机气路系统故障诊断进行了深入研究。针对某型真实飞机发动机进行测试试验采集的气路多参量数据,首先采用经验模态分解(Empirical Mode Decomposition,EMD)对气路系统各参量信号进行EMD分解,然后采用软阈值函数对其进行降噪,并进行信号重构,从而可得到飞机发动机气路工作状态有效数据。在此基础上,设计了飞机发动机气路系统主元分析故障诊断模型,并结合预处理得到的飞机发动机气路有效数据,运用所设计的主元分析故障诊断模型对飞机发动机进行故障诊断技术研究。研究结果表明,所提出的方法能够很好地诊断出飞机发动机气路系统实际运行时所出现的故障,具有重要的实际应用价值,并有广泛的应用前景。     

A characterization of optimal base‐stock levels for a multistage serial supply chain

In this article, we present a multistage model to optimize inventory control decisions under stochastic demand and continuous review. We first formulate the general problem for continuous stages and use a decomposition solution approach: since it is never optimal to let orders cross, the general problem can be broken into a set of single‐unit subproblems that can be solved in a sequential fashion. These subproblems are optimal control problems for which a differential equation must be solved. This can be done easily by recursively identifying coefficients and performing a line search. The methodology is then extended to a discrete number of stages and allows us to compute the optimal solution in an efficient manner, with a competitive complexity. © 2016 Wiley Periodicals, Inc. Naval Research Logistics 63: 32–46, 2016     

The probability of the existence of a feasible flow in a stochastic transportation network

Stochastic transportation networks arise in various real world applications, for which the probability of the existence of a feasible flow is regarded as an important performance measure. Although the necessary and sufficient condition for the existence of a feasible flow represented by an exponential number of inequalities is a well‐known result in the literature, the computation of the probability of all such inequalities being satisfied jointly is a daunting challenge. The state‐of‐the‐art approach of Prékopa and Boros, Operat Res 39 (1991) 119–129 approximates this probability by giving its lower and upper bounds using a two‐part procedure. The first part eliminates all redundant inequalities and the second gives the lower and upper bounds of the probability by solving two well‐defined linear programs with the inputs obtained from the first part. Unfortunately, the first part may still leave many non‐redundant inequalities. In this case, it would be very time consuming to compute the inputs for the second part even for small‐sized networks. In this paper, we first present a model that can be used to eliminate all redundant inequalities and give the corresponding computational results for the same numerical examples used in Prékopa and Boros, Operat Res 39 (1991) 119–129. We also show how to improve the lower and upper bounds of the probability using the multitree and hypermultitree, respectively. Furthermore, we propose an exact solution approach based on the state space decomposition to compute the probability. We derive a feasible state from a state space and then decompose the space into several disjoint subspaces iteratively. The probability is equal to the sum of the probabilities in these subspaces. We use the 8‐node and 15‐node network examples in Prékopa and Boros, Operat Res 39 (1991) 119–129 and the Sioux‐Falls network with 24 nodes to show that the space decomposition algorithm can obtain the exact probability of these classical examples efficiently. © 2016 Wiley Periodicals, Inc. Naval Research Logistics 63: 479–491, 2016     

On the hierarchy of γ‐valid cuts in global optimization

Concavity Cuts play an important role in concave minimization. In Porembski, J Global Optim 15 ( 17 ), 371–404 we extended the concept underlying concavity cuts which led to the development of decomposition cuts. In numerical experiments with pure cutting plane algorithms for concave minimization, decomposition cuts have been shown to be superior to concavity cuts. However, three points remained open. First, how to derive decomposition cuts in the degenerate case. Second, how to ensure dominance of decomposition cuts over concavity cuts. Third, how to ensure the finite convergence of a pure cutting plane algorithm solely by decomposition cuts. These points will be addressed in this paper. © 2007 Wiley Periodicals, Inc. Naval Research Logistics, 2008     

捷联惯导大失准角误差模型在快速传递对准中的应用

针对快速传递对准中主子惯导相对姿态存在大角度的情况,推导了捷联惯导大失准角误差模型.该模型采用欧拉角表示姿态误差,并用欧拉运动方程准确描述其传播规律.鉴于该模型中的姿态观测方程是复杂的非线性函数,采用无需求导的UKF算法,并采用奇异值分解(SVD)解决方差阵的病态问题.仿真结果表明,该算法在小角度误差条件下滤波精度优于线性模型,并且适用于大角度误差条件.