The back - propagation algorithm also rests on the idea of gradient descent , and so the only change in the analysis of weight modification concerns the difference between t and y 反向传播算法同样来源于梯度降落原理，在权系数调整分析中的唯一不同是涉及到t ( p , n )与y ( p , n )的差分。
The system denotes knowledge with status space , manages repository and correlative database by use of dbms , and speculate mechanism makes use of backward promulgate algorithm of three layer artificial neural network 该系统采用状态空间知识表示方法，应用dbms管理知识库和相关数据库，推理机采用三层人工神经网络的反向传播算法。
Some developed bp algorithms which were simulated and compared using computer seperately are also presented , ann learning algorithm in this paper is constructed on the basis of the goodness of all above algorithms which is suitable for the identification of typical fault mode 同时，对神经网络的主要学习算法之一的反向传播算法( bp算法)进行了深入的研究，通过对几种改进算法的模拟实现和对比综合，构造了适合于对箱式变电站典型故障进行识别的学习算法。
As for turbine , the importance of its fault diagnosis is introduced ; the research status quo in this field all over the world and the difference from variable diagnosis method is presented . hybrid model fault diagnosis method based on expert system and back - propagation algorithm is applied to solve the turbine fault diagnosis . good effect is obtained 针对汽轮机故障诊断问题，本论文剖析了汽轮机故障诊断的重要性，论述了国内外相关领域的研究现状，分析了多种诊断方法的优缺点，选择了基于专家系统和基于反向传播算法的混合模型的故障诊断方法来解决该问题，取得了很好的效果，全文的主要研究内容和成果总结如下： ( 1 )概述了一些现有的故障诊断方法。
This article is established in fault diagnosis research of automatic test , makes discuss and study about some pivotal techniques in software design of automatic test and fault diagnosis , including report forms printing , backward promulgate algorithm , database management , and etc . and also advances the author ' s experience and understanding during debugging 本文立足于基于自动测试的故障诊断系统的研制工作，对自动测试和故障诊断的软件设计中的几个关键技术，包括报表打印、反向传播算法和数据库管理等进行了讨论和研究，并提出了作者在实际工作过程中的一些体会和理解。
Firstly , the square character of pd gray degree images is studied , which is used in describing geometry character of pd gray degree images ; secondly , the relative statistic character parameters of pd gray degree images is studied . by calculating statistic character parameters of discharge samples , we can also find certain differences between fractal features of different discharge exist . ( 4 ) te classifier with back - propagation neural network ( bpnn ) is designed 首先研究了局部放电灰度图像矩特征，利用矩特征描述局部放电灰度图像基本几何形状特征；其次研究了局部放电灰度图像的相关统计特征参数；通过对放电样本的统计特征参数计算，结果表明不同类型放电的统计特征具有一定的差异； ( 4 )设计了反向传播算法前馈人工神经网络分类器( backpropagationneuralnetwork ，简称为bpnn ) ，研究了在采用上述识别特征和bpnn识别局部放电灰度图像的方法；识别结果表明，该方法能够有效地应用于局部放电模式自动识别，并获得了良好的识别效果。
Based on experimental data , the bp model is built , which reflects the relationship between the mechanical properties , phase transformation points of material and its composition , micro - structure . hardenability curves , alloy composition , austenite temperature , martensite start temperature and yield strength are predicted by means of artificial neural network technology 本研究在实验数据的基础上，采用反向传播算法( bp算法)建立了反映合金成分-组织-性能之间的定量关系的网络模型，并利用所得到的模型，分别对淬透性曲线、合金成分、力学性能、相变点进行分析和预测。
The method has special obvious advantages to diagnosis the faults when several faults exist simultaneously . the paper constructs a three - layer forward neural network to diagnosis the fault and trains the network with characteristic eigenvectors extracted through the wavelet packet analysis , when training the net using the method of adding inertia item and the bp algorithm of conjugate gradient method , at the same time adapt the initial data of the network through the test 同时，文中还构造了三层的前向神经网络，以小波包分析提取的故障特征向量作为网络的训练样本数据，采用加惯性因子、共轭梯度法和迭代过程中改变学习率的反向传播算法来对神经网络进行训练，并采用试验的方法调整神经网络的初始值。
Firstly , on the basic of normal error back propagation algorithm ( bp a1gorithm ) , the model was added the inertia impulse item in the updating formula for weight , and then let learning rate and inertia parameter adjust ll self - adaptively , so the improved bp algorithm ( improved bp algorithm , ibp algorithm ) formed 首先，该模型在误差反向传播算法( bp算法)的基础上，在网络权值更新公式中添加惯性冲量项，并对学习率和惯量因子进行自适应调整，从而形成bp改进算法( improvedbp算法，简称ibp算法) 。
