网络函数

Another carries out input fuzzy quantificat and grading quantificat in the same time, deduces using Takagi-Sugeno fuzzy algorithm, and introduces fuzzy algorithm into weight training of CMAC. It has the characteristics of less quantificat grade and high accuracy of function approximability. It shows good effectivity used for Wiener model identifying. By syncretizing MVL, linear neuron, fuzzy control and CMAC, this paper gives a kind of FCMAC nonlinear control system with the ability of self-learning in line.

将模糊控制和小脑模型神经网络有机的结合在一起，提出了两种模糊小脑模型神经网络，一种是在CMAC基础上进行输出模糊插值，可减小非样本点误差，提高精度和训练速度；一种是对输入进行分级量化的同时进行模糊量化，利用Takagi-Sugeno模糊算法进行推理，并将模糊算法引入CMAC的权值训练，具有输入量化级数少、函数逼近精度高等特点，用于Wiener模型辨识效果好。

Nonlinear function approximation using BP neural network and RBF neural network is simulated.

最后，作了用BP网络和RBF网络逼近非线性函数的仿真研究，并作了比较。

The zonal translation networks on the unit sphere is constructed with the best approximation of spherical harmonic polynomials and the Riesz means. The Jackson theorem of approximation by such kind of zonal translation is established.

借助于球调和多项式的最佳逼近多项式和Riesz平均构造出了单位球面S上的带形平移网络，并建立了球面带形平移网络对LS(上标 q中函数一致逼近的Jackson型定理。

In Chapter 2, by using M-matrix theory, some analysis techniques and constructing suitable Lyapunov functionals, we investigate the convergence of solutions and the existence and uniqueness of periodic solution for system:where ft and g, are the propagational signal functions defined onR.

在第二章中，讨论一类推广的具变时滞和变系数的双向联想记忆神经网络模型的渐近状态。这里，信号传输函数f_i，g_i(i=1，2，…，n，j=1，2，…，m)是R→R上的连续函数，且满足Lipschitz条件。

At first, this paper gives the brief introduction of the history and present status of accelerometer and its testing technology, the working principium and math model of the liquid floated pendu-lous accelerometer, and then, decides the binary width pulse force retrim loop as the design proposal of testing system, researches the transfer function of every part in the system emphasizly, analyses the stability of the whole accelerometer testing system from the angle of control theoretics by the open loop transfer function of system, and designed the correcting net, analyses the basal problems such as resolution, sampling restraint, precision and so on, designs the hardware testing circuits such as preamplification, band-pass filter, alternating amplifier, phase sensitive demodulatorn, pulse-width modulation, frequency scale circuit, moment current generator. Finally, using the graphics program language LabV-IEW which is designed for testing field especially by NI accomplishes the solfware design of testing system,realized the testing functions.

首先对加速度计及其测试技术的发展历史和现状，液浮摆式加速度计的工作原理和数学模型等作了简要的介绍，然后确定了以二元调宽脉冲再平衡测试回路为设计方案，并从控制理论的角度进行了分析，着重研究了系统中各部分的传递函数，利用系统开环传递函数分析了系统的稳定性，同时设计了系统的校正网络；分析了二元调宽脉冲再平衡测试回路的分辨率、采样约束以及测试精度等基本问题，并按照系统分析的结果设计了包括前置放大、带通滤波、交流放大、相敏解调、脉宽调制、频标电路以及力矩电流发生器等测试系统各部分硬件电路，验证了电路的正确性，最后按照测试系统的要求，采用了美国NI公司专为测试领域所开发的虚拟仪器工具——LabVIEW作为测试软件开发工具，利用该图形化编程语言完成了测试系统软件部分的设计，实现了测试功能。

Methods This paper set up the functional projective relationship from the imput to the output ,utilizing the function approximating model of artificial neural network.

利用人工神经网络的函数逼近功能模型，建立输入到输出的函数映射关系。

The thesis proposes a latency jittering adaptive controlling algorithm with the end-value forecasting function. Besides automatically adjusting the latency jittering value, the algorithm can also automatically adjust the value of TokenRate, PeakBandwidth and other QoS values.

提出了一种基于IP网络的优先级调度，自适应QoS收缩算法，自适应QoS协商机制和自适应QoS接入控制流程；基本思想是基于RSVP，在提出QoS请求时进行QoS映射，然后启动适应性函数和资源管理函数进行协商，直到获得一组在当前资源状况下最佳的QoS指标。

A new method for fast valving decision to maintain multi-machine power systems transient stability by feedforward neural network is proposed.

这种输入和输出的选取是通过详细的理论分析，使输入和输出间呈连续函数关系，再经对输入量进行简化而决定的，从而应用前向神经网络的函数逼近功能说明了所提神经网络结构的合理性。

Some global properties such as global at tractivity and global exponential stability for delayed Hopfield neural networks model, under the weaker assumptions on nonlinear activation functions, are conc erned.

对具有时滞的Hopfield神经网络模型，在非线性神经元激励函数是Lipschitz连续(而非已有的大部分文献中假设是Sigmoid函数)的条件下，通过构造适当的泛函，给出了这类模型全局吸引和平衡点全局指数稳定的易于验证的充分条件。

On the basis of both adaptive BP algorithm and Newton's method, Quasi Newton algorithm with adaptive decoupled step and momentum for feed-forward neural networks is derived.

基于输出层函数为线性函数的三层前馈神经网络，结合自适应步长和动量解耦的伪牛顿算法及迭代最小二乘法导出了一种混合算法。

Neither the killing of Mr Zarqawi nor any breakthrough on the political front will stop the insurgency and the fratricidal murders in their tracks.

在对危险的南部地区访问时，他斥责什叶派民兵领导人对中央集权的挑衅行为。

In fact,I've got him on the satellite mobile right now.

实际上 我们已接通卫星可视电话了