基于DEMON谱和LSTM网络的水下运动目标噪声基频检测*

传统的船舶辐射噪声基频检测方法不仅依赖大量的先验知识,而且对背景噪声非常敏感。为了提高目标识别的稳定性和精确性,本文提出了一种基于深度神经网络的基频检测算法。首先从多通道水听器信号中提取DEMON谱,然后直接将二维谱特征矩阵输入由CNN和LSTM构成的级联网络,最后通过稠密层输出实现对基频的估计。从仿真和外场试验数据得到如下结论：（1）深度网络能够实现无先验知识和不同信噪比条件下的基频检测,具有良好的泛化性能。（2）LSTM网络能够高效地从时序DEMON谱中提取统计特征,提高基频估计精度。（3）输入信号的时间长短会影响网络的检测精度,更长时间的信号能够获得更好的检测结果。     

基于L型延迟阵列调制宽带转换器的信号载频和二维到达角联合估计

针对现有的基于欠采样的频率和二维到达角的联合估计存在结构复杂问题,本文提出了一种基于调制宽带转换器技术的L型延迟阵列接收结构.利用延迟通道与未延迟通道采样值之间的相位差可直接估计载频,进而计算二维到达角,无需额外的参数配对操作,避免了配对步骤引入的误差和复杂度的提升.并结合所提L型延迟阵列结构的特点构造相关矩阵和三线性模型,提出了两种参数估计算法,一种基于旋转不变子空间算法,计算量小,适用于需要实时处理的场景;另一种基于正则分解技术,鲁棒性较好,适用于信噪比较低的应用场景.仿真实验表明该方法能较好地从欠奈奎斯特样本中估计目标的载频和二维到达角参数.     

光纤相控阵稀疏排布优化算法对比

光纤相控阵在激光合束、激光雷达等领域具有应用前景.光纤阵列配置方式不同于微波相控阵,光纤天线间距大于波长时存在周期旁瓣问题,影响主瓣能量分布.本文从物理模型出发,建立了基于同心圆环形点阵集合的光学相控阵天线布阵理论模型,提出了利用解析延拓的傅里叶变换方法实现干涉场强度的快速合成理论,讨论了在离散采样时数值仿真需关注的采样带宽和采样数目问题,解决了快速实现多光束干涉场数值仿真的问题.对比研究了两种优化光学相控阵天线配置的优化算法:遗传算法和粒子群算法,分别实现了不同种群数量遗传算法和粒子群算法迭代优化,分析了二者在优化过程中的收敛速度和优化效果,得到了峰值旁瓣比PSR=0.270的配置阵列.所提出的方法有望用于实际的光学相控阵天线配置中,指导天线主瓣能量最大化的优化设计;研究模型对不可微分目标函数优化问题的研究有一定参考价值.     

定性和定量分析磺胺氯吡嗪钠可溶性粉中的未知添加物

定性和定量分析了一批兽药磺胺氯吡嗪钠可溶性粉中的未知添加物。使用电位滴定法对该批磺胺氯吡嗪钠可溶性粉进行含量测定,发现滴定异常,根据标准进行的两项鉴别有一项不合格,因此怀疑添加了其他化合物。采用超高效液相色谱-四极杆-飞行时间质谱(UHPLC-Q/TOF MS)对该批样品进行筛查,发现疑似添加物,并使用高效液相色谱-二极管阵列检测(HPLC-DAD)法进行了双重确证和含量测定。该批样品中非法添加物为乙酰甲喹和磺胺二甲嘧啶,添加量分别为40.3和16.4 mg/g。通过总结该批样品非法添加物的发现、确证和含量测定的整个过程,得出一种较好的筛查未知物的模式,为兽药处方外非法添加筛查提供可借鉴的思路。     

Computation methods for combinatorial sums and Euler‐type numbers related to new families of numbers

The aim of this article is to define some new families of the special numbers. These numbers provide some further motivation for computation of combinatorial sums involving binomial coefficients and the Euler kind numbers of negative order. We can show that these numbers are related to the well‐known numbers and polynomials such as the Stirling numbers of the second kind and the central factorial numbers, the array polynomials, the rook numbers and polynomials, the Bernstein basis functions and others. In order to derive our new identities and relations for these numbers, we use a technique including the generating functions and functional equations. Finally, we give not only a computational algorithm for these numbers but also some numerical values of these numbers and the Euler numbers of negative order with tables. We also give some combinatorial interpretations of our new numbers. Copyright © 2016 John Wiley & Sons, Ltd.     

基于声矢量水听器阵列误差的自校正算法

对声矢量水听器阵列的各类误差进行了分类,推导了各类误差对阵列信号模型的影响因子,通过Monte Carlo实验分析对比了各类误差对阵列DOA估计性能的影响,然后将方向性误差和位置误差归结为幅度误差和相位误差,在传统声压阵列误差校正模型和算法的基础上,得到矢量阵列误差自校正的优化模型及自校正算法,最后,通过仿真实验和外场实验的数据处理表明,自校正算法具有良好的参数估计性能,具有一定的工程实用性.     

On the restricted Chebyshev–Boubaker polynomials

Using the language of Riordan arrays, we study a one-parameter family of orthogonal polynomials that we call the restricted Chebyshev–Boubaker polynomials. We characterize these polynomials in terms of the three term recurrences that they satisfy, and we study certain central sequences defined by their coefficient arrays. We give an integral representation for their moments, and we show that the Hankel transforms of these moments have a simple form. We show that the (sequence) Hankel transform of the row sums of the corresponding moment matrix is defined by a family of polynomials closely related to the Chebyshev polynomials of the second kind, and that these row sums are in fact the moments of another family of orthogonal polynomials.     

Application of sequential factorial design and orthogonal array composite design (OACD) to study combination of 5 prostate cancer drugs

Prostate cancer is one of the most common cancers among men in the United States. It is also a major leading cause of cancer death among men of all races. In order to treat prostate cancer, drug combinations are often applied. Drug combinations target at different pathways of cells can potentially lead to higher efficacy and lower toxicity due to drug synergy. In this paper, we sequentially applied a two-level design and a follow-up orthogonal array composite design (OACD) to investigate combinations of five anti-cancer drugs, namely, doxorubicin, docetaxel, paclitaxel, cis-dichlorodiamine platinum and dihydroartemisinin. Our initial screening using a two-level full factorial design identified doxorubicin and docetaxel as the most significant drugs. A follow-up experiment with an OACD revealed more complicated drug interactions among these 5 anti-cancer drugs. Quadratic effects of doxorubicin and paclitaxel appeared to be significant. A further investigation on contour plots of all the two-drug pairs indicated that combination of doxorubicin and docetaxel are the most effective companion, while the combination of cis-dichlorodiamine platinum and dihydroartemisinin showed unknown antagonistic effects which diminished the individual drug anti-cancer efficacy. These observations have significant practical implications in the understanding of anti-cancer drug mechanism that can facilitate clinical practice of better drug combinations.     

A transmit/receive volume strip array and its mode mixing theory in MRI

MRI is increasingly moving towards higher magnetic field, prompting the need for multi-port transmit/receive RF coil arrays to overcome high-frequency limitations such as penetration depth and dielectric resonance effects. In this work, an arbitrary n-element transmit/receive volume strip array (VSA) and an associated mixing mode theory are described to understand the behavior of a multiple-port cyclic symmetrical VSA in both the physical port space and the complementary mode space; the relations between the two spaces are explicitly formulated. The advantage of mode-space analysis is that an arbitrary n x n impedance matrix which describes any VSA in port space can be diagonalized to a diagonal n x n matrix; thus an analytical solution of Kirchhoff's laws for the VSA becomes manageable when n is large. Based on such an analytical solution, we can (a) generate excitation profile of any desired mixed mode during transmission by manipulating external power sources without the need of physically tuning VSA to the mixed mode; (b) identify the sensitivity profiles of the complementary mode distributions during reception, which were unknown in quadrature and decoupled coils. Many predictions are rigidly verified by extensive test measurements from network analyzer and by MR imaging experiments.     

Signal-to-noise ratio comparison of phased-array vs. implantable coil for rat spinal cord MRI

The signal-to-noise ratio (SNR) performance and practicality issues of a four-element phased-array coil and an implantable coil system were compared for rat spinal cord magnetic resonance imaging (MRI) at 7 T. MRI scans of the rat spinal cord at T10 were acquired from eight rats over a 3 week period using both coil systems, with and without laminectomy. The results demonstrate that both the phased array and the implantable coil systems are feasible options for rat spinal cord imaging at 7 T, with both systems providing adequate SNR for 100-mum spatial resolution at reasonable imaging times. The implantable coils provided significantly higher SNR, as compared to the phased array (average SNR gain of 5.3x between the laminectomy groups and 2.5x between the nonlaminectomy groups). The implantable coil system should be used if maximal SNR is critical, whereas the phased array is a good choice for its ease of use and lesser invasiveness.