神经网络

This paper introduces the homotopy algorithm into back-propagation neural network and indicates that the convergence speed can be improved and the defects of converging to local extremum can be eliminated.

同伦算法是解决非线性优化问题的新理论，将同伦算法引入误差反向传播神经网络，提高了网络的训练速度，消除了常规BP网络算法收敛于局部极值的缺点。

Second, the conventional linear optimal control can make system stabilize at neighborhood of balance point only. In this paper, we introduced a method of control by Artificial Neural Network. And also, identified variable on line. This method enhanced the adaptability of the control.

其次，由于常规的线性最优控制只能在平衡点附近稳定，这里利用人工神经网络在线识别不同运行状况下的各个参数值，离线训练网络，最后用训练成功的网络进行在线控制，实现了系统的大范围稳定。

A method is proposed in order to accelerate the convergence of artifcial neural network modeling of chemical process,in which the principal components analysis is applied in the data processing in training sets.

文摘：达到提出应用主成分分析法对样本进行预处理，减少网络的输入因子数，消除输入因子间的相关性并简化网络结构，达到提高网络学习速率的目的，得到的人工神经网络模型能达到所要求的精度。

It is presented to choose σ parameter of kernel function on dynamic, which enhances preciseness rate of diagnosis.

神经网络方法(Neural Networks，NN)在故障诊断中的应用研究取得了很多的成果，但它在网络结构选择、网络的训练及提高网络的泛化能力等方面存在的许多问题，限制了它在工程实际中的广泛应用。

Combined chaos theory with Neural network, it presents predisposal technology in data-disposal of phase space, compares it with others, its convergent speed and forecasting precision have been improved greatly.

将混沌理论与神经元网络结合起来，首次提出了在相空间中进行数据预处理的神经网络训练样本预处理技术，该技术通过与其他预处理技术的比较，其网络收敛速度与预测精度有较大提高。

First, the paper introduces the sturcture of feedforward neural network and classical back-propagation learning algorithm, then systematic analysis and reserch are made to the various learning methods of RBFNN(Radical Basis Function Neural Network). The idea and algorithm frame of evolution strategy are presented and introduced to the learning algorithm of BP Neural network to improve the performance of it.

本文首先阐述了前向神经网络的结构以及经典的反向传播学习算法，尤其是对RBF网络的各种学习算法进行了比较系统的分析研究，然后又介绍了进化计算的思想和算法框架，并把这种算法思想引入到BP网络学习思想中去对BP算法进行了改进。

A new hypnotizing/waking technique of net node is proposed to reduce the scale of linear parameter neural network.

对于线性参数神经网络实现时规模庞大的问题，提出了网络节点的催眠/唤醒技术，使动态网络的规模降低到最小。

In this paper, a large number of simulations have been done based on the two different networks. Simulations show that the transformer fault diagnosis system based on the probability neural network is better than BP network system in the speed, accuracy and the ability for samples superaddition.

本文对基于不同网络模型的两个系统进行了大量仿真，仿真结果表明基于概率神经网络的变压器故障诊断系统在速度、正确率以及追加样本能力等各方面的能力都要优于BP网络系统。

To overcome the disadvantages of standard BP networks, two unproved algorithm are put forward. One is information entropy based, and the other is genetic algorithm based.

针对标准BP神经网络的缺点，提出了两种改进方法：①基于信息优化的网络学习算法；②基于遗传算法的网络学习算法。

A robust output feedback controller is advanced by using the DRNN-based observer presented above, so that the output-matching and lipschitz conditions are not imposed on the uncertainties; Robust adaptive controller is used in a class of nonlinear systems based on neural networks and H〓 methods, thereby the constraint on uncertainties with largest known boundness is removed, not as in H〓 methods.

利用本文的动态神经网络鲁棒观测器，设计了鲁棒输出反馈控制器，取消了已有结果对未知动态的输出匹配条件的限制；结合静态神经网和H〓控制方法，对一类非线性系统设计了鲁棒自适应反馈控制器，克服了H〓控制要求不确定性最大可能界已知的不足，其最大优点是仅要求网络逼近误差有界，且不要求网络权的离线训练。

We have no common name for a mime of Sophron or Xenarchus and a Socratic Conversation; and we should still be without one even if the imitation in the two instances were in trimeters or elegiacs or some other kind of verse--though it is the way with people to tack on 'poet' to the name of a metre, and talk of elegiac-poets and epic-poets, thinking that they call them poets not by reason of the imitative nature of their work, but indiscriminately by reason of the metre they write in.

索夫农 、森那库斯和苏格拉底式的对话采用的模仿没有一个公共的名称；三音步诗、挽歌体或其他类型的诗的模仿也没有——人们把&诗人&这一名词和格律名称结合到一起，称之为挽歌体诗人或者史诗诗人，他们被称为诗人，似乎只是因为遵守格律写作，而非他们作品的模仿本质。

The relationship between communicative competence and grammar teaching should be that of the ends and the means.

交际能力和语法的关系应该是目标与途径的关系。