具有算子约束的Lipschitz规划的最优性条件

Barbu等人在文[1]中时目标函数和约束算子都是Frechet可微的情况下证明了具有算子约束的数学规划的最优性必要条件．本文将这一问题推广为目标函数为非光滑的情形,给出了具有算子约束的Lipschitz规划的最优性充分条件和必要条件．     

具有增益的概率网络金融计划模型

本文绘出一类具有增益的概率网络金融计划模型．许多多阶段金融计划问题可纳入这类模型．在这类模型中,随机变量的分布函数与Alexander过滤交易规则密切联系在一起,金融市场交易信号由神经网络产生,目标函数的最优值按其期望值计算．文中提出临界流和临界路的概念,给出目标函数下界等于其期望值的充分必要条件和期望最优解的求解方法．     

An In4(Se3)1−x Te3x solid solution as a material for absorption optical filters

We suggest using crystals of the In₄(Se₃)_1−xTe_3x solid solution for manufacturing absorption optical filters with the boundary wavelength regulated from 1.7 to 2.5 μm by changing the composition of the solution. The refractive index is determined for different compositions, and the possibility of making the filters antireflecting is shown. The advantages of the proposed filters are connected with the specific features of the absorption edge of crystals and their structure. To whom correspondence should be addressed. Chernovtsy State University, 2, Kotsyubinskii Str., Chernovtsy, 274012, Ukraine. Translated from Zhurnal Prikladnoi Spektroskopii, Vol. 66, No. 4, pp. 577–579, July–August, 1999.     

纯位相信息自聚焦计算全息匹配滤波器

本文介绍一种用计算全息方法制作的纯位相信息自聚焦匹配滤波器。它在编码时取滤波函数的振幅为１,仅对位相进行编码;并通过对傅里叶谱加入二次位相因子代替傅里叶透镜,使其具有自聚焦功能。与常规的方法（光学全息法或计算全息法）所获得的匹配滤波器相比较,具有原理上更合理,滤波器的制作和匹配光路更简单,也更容易实现匹配等特点。所得到的匹配实验结果也很清晰,这与理论是完全符合的。     

The Resonance Peak Theory of Reflection Guidedmode Resonance Filters

Ｉｎｔｒｏｄｕｃｔｉｏｎ　　Ｉｎ１９６５,ＨｅｓｓｅｌａｎｄＯｌｉｎｅｒ［１］,ｂｙｕｓｉｎｇｒｉｇｏｒｏｕｓｅｌｅｃｔｒｏｍａｇｎｅｔｉｃｔｈｅｏｒｙ,ａｎａｌｙｚｅｄｔｈｅｒｅｆｌｅｃｔｉｖｅｐｒｏｐｅｒｔｉｅｓｏｆａｓｉｎｕｓｏｉｄａｌｒｅｆｒａｃｔｉｏｎｇｒａｔｉｎｇｓ（ｉ．ｅ．,ｇｅｎｅｒａｌｈｏｌｏｇｒａｐｈｉｃｇｒａｔｉｎｇｓ,ｗｈｉｃｈｗｉｔｈｅｑｕａｌｔｈｉｃｋｎｅｓｓａｎｄｓｉｎｕｓｏｉｄａｌｍｏｄｕｌａｔｉｏｎｏｆｒｅｆｒａｃｔｉｖｅｉｎｄｅｘａｌｏｎｇｔｈｅｓｕｒｆａｃｅ…     

Suppression for the Low spatial frequency Rings and Ripples of Laser Near field Intensity Distribution in the Spatial Filter Used in ICF Laser DriveSupported by the key project of the National Science Foundation, National Hi Tech Foundation.

１ＩｎｔｒｏｄｕｃｔｉｏｎＵｎｉｆｏｒｍｉｔｙｏｆｔｈｅｆｌｕｅｎｃｅｄｉｓｔｒｉｂｕｔｉｏｎｄｅｌｉｖｅｒｅｄｂｙａｌａｓｅｒｉｓｅｓｅｎｔｉａｌｆｏｒｍａｎｙａｐｐｌｉｃａｔｉｏｎｓ,ｓｕｃｈａｓｏｐｔｉｃａｌｐｒｏｃｅｓｉｎｇ,ｓｕｒｆａｃｅｔ...     

A Study of Dipolar Interactions and Dynamic Processes of Water Molecules in Tendon byH andH Homonuclear and Heteronuclear Multiple-Quantum-Filtered NMR Spectroscopy

The effect of proton exchange on the measurement of¹H–¹H,¹H–²H, and²H–²H residual dipolar interactions in water molecules in bovine Achilles tendons was investigated using double-quantum-filtered (DQF) NMR and new pulse sequences based on heteronuclear and homonuclear multiple-quantum filtering (MQF). Derivation of theoretical expressions for these techniques allowed evaluation of the¹H–¹H and¹H–²H residual dipolar interactions and the proton exchange rate at a temperature of 24°C and above, where no dipolar splitting is evident. The values obtained for these parameters at 24°C were 300 and 50 Hz and 3000 s⁻¹, respectively. The results for the residual dipolar interactions were verified by repeating the above measurements at a temperature of 1.5°C, where the spectra of the H₂O molecules were well resolved, so that the¹H–¹H dipolar interaction could be determined directly from the observed splitting. Analysis of the MQF experiments at 1.5°C, where the proton exchange was in the intermediate regime for the¹H–²H dipolar interaction, confirmed the result obtained at 24°C for this interaction. A strong dependence of the intensities of the MQF signals on the proton exchange rate, in the intermediate and the fast exchange regimes, was observed and theoretically interpreted. This leads to the conclusion that the MQF techniques are mostly useful for tissues where the residual dipolar interaction is not significantly smaller than the proton exchange rate. Dependence of the relaxation times and signal intensities of the MQF experiments on the orientation of the tendon with respect to the magnetic field was observed and analyzed. One of the results of the theoretical analysis is that, in the fast exchange regime, the signal decay rates in the MQF experiments as well as in the spin echo or CPMG pulse sequences (T₂) depend on the orientation as the square of the second-rank Legendre polynomial.     

Fabrication and Performance of 1.55 μ/m Transmission Glass-Echelon Grating Filters

This paper describes the fabrication method and the measured spectral behavior of a 1.55 μm transmission glass-echelon grating filter. This echelon grating consists of a pile of identical plane-parallel glass plates constructed using a special assembling tool. The experimental result confirmed the accuracy of the Fourier transform which indicated that the spectral passband width was inversely proportional to the beam size. This feature makes it possible to use the filter as a bandpass filter or a comb filter when constructing photonic networks. The echelon filter has certain intrinsic advantages including a low insertion loss and polarization independence. Moreover, it offers good cost performance and excellent design flexibility.     

Spectral Analysis of Transmission Echelon Grating Filters for Photonic Networks

This paper describes a method of analyzing spectral behavior that allows transmission echelon grating filters to be designed for use in constructing photonic networks. These networks include dense wavelength division multiplexing systems and other such applications. First, the configuration and principle of this filter are described. Then, the transmittance spectrum of the transmission echelon grating filter is considered for incident lights with rectangular, Gaussian, and sinc electric field distributions. Next, a more realistic spectral response is considered by taking reasonable fabrication errors into account. Finally, the group delay and the transmittance of linearly chirped echelon grating filters are analyzed to access their potential for use as a wavelength dispersion equalizer. These analytical results make it possible to design practical echelon grating filters.     

奇异值函数的二阶方向导数

本文用一个直接的方法给出了奇异值函数的二阶方向导数公式. 作为应用, 利用这一公式建立了谱范数的上图集合与核范数的上图集合的切锥和二阶切集的具体表达式, 这些表达式在矩阵优化的一阶和二阶最优条件的研究中起着重要作用.