高密度封装CAD技术研究

介绍了自行研制的高密度封装外壳设计软件、阐述了其工作原理和功能。应用该软件设计并制作出目前国内最多针脚数的外壳ＰＧＡ２５７。     

基于邻域统计特性的概率神经网络及其在自动目标识别中的应用

在概率神经的一种改进模型－ＦＤＯ网络的基础上，提出在设计网络收敛域时进一步考虑每一像素点周围８邻域的影响，对网络的作用函数加以修，使改进后的网络具有稳定性好且收敛速度快的优点。通过实验对改进前后网络的识别性能加以比较，证明改进后的网络特别适用于噪声图像的识别。     

Looking at Fokker-Planck dynamics with a noisy instrument

We consider the class of experiments which can be characterized by a Fokker-PIanck dynamics corresponding to the overdamped motion of a state point in a suitable stochastic potential. We assume that the general form of the potential is known (or can be guessed with reasonable accuracy), but that its parameters are to be determined experimentally by measurements made with a noisy instrument. This possible method for determining the potential parameters, which exploits the system's own internal stochastic motion in order to explore rapidly its available parameter space, is substantially more efficient than traditional methods involving time averages of single point measurements, and yet does not appear to have been previously considered. The method could be important when, for example, the experiment must be completed in a limited time owing either to the expense of the experimental materials or to the temporary stationarity of the preparation, situations which are commonly encountered in experimental biochemistry and biology.     

On the phase space approach to complexity

We elaborate in some detail on a new phase space approach to complexity, due to Y.-C. Zhang. We show in particular that the connection between maximal complexity and power law noise or correlations can be derived from a simple variational principle. For a 1D signal we find 1/f noise, in accordance with Zhang.     

Model-dependent nature of blowtorch effect on the escape and equilibration rates

We analyze the relaxation behavior of a bistable system when the background temperature profile is inhomogeneous due to the presence of a localized hot region (blowtorch) on one side of the potential barrier. Since the diffusion equation for inhomogeneous medium is model-dependent, we consider two physical models to study the kinetics of such system. Using a conventional stochastic method, we obtain the escape and equilibration rates of the system for the two physical models. For both models, we find that the hot region enhances the escape rate from the well where it is placed while it retards the escape rate from the other well. However, the value of the escape rate from the well where the hot region is placed differs for the two models while that of the escape rate from the other well is identical for both. This work, for the first time, gives a detailed report of the similarities and differences of the escape rates and, hence, exposes the common and distinct features of the two known physical models in determining the way the bistable system relaxes. Received 25 September 2001     

电子直线加速器磁控管频率稳定的自适应线性神经元方法

 电子直线加速器电子束能量的稳定与否取决于功率源工作频率的稳定性,磁控管在短时间内的散谱和轻微跳谱造成稳频系统的控制精度下降,最后电子束的扫描均匀度下降。引入自适应线性神经元方法(ADALINE)和噪声对消技术以消除对工作频率长期稳定性的影响,从而保证了电子束的扫描均匀度。     

Quasipotentials for simple noisy maps with complicated dynamics

The theory of nonequilibrium potentials or quasipotentials is a physically motivated approach to small random perturbations of dynamical systems, leading to exponential estimates of invariant probabilities and mean first exit times. In the present article we develop the mathematical foundation of this theory for discrete-time systems, following and extending the work of Freidlin and Wentzell, and Kifer. We discuss strategies for calculating and estimating quasipotentials and show their application to one-dimensionalS-unimodal maps. The method proves to be especially suited for describing the noise scaling behavior of invariant probabilities, e.g., for the map occurring as the limit of the Feigenbaum period-doubling sequence. We show that the method allows statements about the scaling behavior in the case of localized noise, too, which does not originally lie within the scope of the quasipotential formalism.     

光外差系统噪声等效功率的一种测量方法

本文介绍了利用工作于Bragg衍射的Ge晶体声光调制器作为外差系统中的频移器来测量外差系统的噪声等效功率(NEP)的方法。该方法具有测试系统简单、信号光功率、偏频可调的优点。实验测试与理论分析吻合。     

一种新的磷化铟复合沟道高电子迁移率晶体管小信号物理模型

针对磷化铟(InP)复合沟道高电子迁移率晶体管(HEMT)的特点，对常规单沟道HEMT的小信号物理模型进行了修正，提出了一种新的用于复合沟道HEMT的小信号物理模型，用商用器件模拟软件ISE(integrated systems engineering)对其进行了仿真验证，对比了实测和仿真的I-V特性及转移特性曲线，重点研究了在InGaAs/InP双层沟道中考虑量子效应后的电场和电流密度随着不同栅电压的变化趋势，研究结果表明，由于在沟道中存在量子效应，在栅下靠源端低电场区域，电流主要分布在InGaAs沟道 关键词： 高电子迁移率晶体管 复合沟道 物理模型 磷化铟     

Comparative Study on Large and Small Signal Theory for Intensity Modulation Response of Semiconductor Lasers Including Higher-Order Dispersion Terms

In this article, the comparison of large signal theory and small signal theory has been done with dispersive propagation of optical signal with IMDD (Intensity Modulation Direct Detection) systems for semiconductor lasers with higher-order dispersion terms. The expressions for an exact large signal theory and small signal theory including higher-order dispersion terms for propagation of an optical wave with sinusoidal amplitude and frequency modulation in a dispersive fiber have been derived. It is observed that small signal theory is more sensitive compared to large signal theory in terms of intensity modulation/direct detection systems. Also, it is reported that for large signal analysis the higher-order effects of dispersion can be ignored, whereas for small signal theory, the higher-order effects can be ignored for lower modulation frequencies only. The variation in the transfer function for various values of modulation indices are greater for small signal analysis than for large signal analysis. Also, as the intensity modulation index is increased, there is a decrease in the value of transfer function. The large signal model approximates the small signal model for lower values of the intensity modulation index.