硬故障

通过构建多类SVM分类器，将间歇故障与正常状态和硬故障分开，以降低BIT虚警。

A multiple SVM classification is designed to distinguish the intermittent failure from the normal and the hard failure .

该文对DHAM建立了同时包含权值误差、权值断路、神经元输出固值等各种软、硬故障的故障模型，定义了概率意义下DHAM的鲁棒性。

DHAM 's mixed fault model was established , including both soft faults such as weights errors and hard faults such as weights broken connections , neuron stuck at faults .

本文在现有文献理论研究的基础上实现了采用BP算法前向多层神经网络对直流测试下模拟电路硬故障的诊断方法。

In the paper , based on the existing literature research foundation an analog circuit catastrophic fault location approach by using feedforward networks with back & propagation learning is realized .

这种新技术首先使用一个小的线性反馈移位寄存器（LFSR），生成伪随机测试模式测试容易测试的故障，并且获得一个硬故障测试立方集T；

Firstly , this new technique uses a small LFSR to generate pseudo random patterns to eliminate the easy to detect faults and obtain a set T of hard fault test cubes .

首先，根据HXD1型机车制动系统的故障特性，将故障分为硬故障和软故障来进行分类研究，给出了硬故障知识诊断和软故障PCA诊断相结合的整体故障诊断方案。

Fault is divided into hard and soft fault to classified research , and a set program of fault diagnosis system is proposed in which hard fault is diagnosed based on knowledge and soft fault is diagnosed based on PCA .

本文成功地完成了硬故障字典的计算机辅助编辑，解决了近似优化及优化技术上的问题。

This paper discusses the optimized Computer-aided Compilation of a hard failure dictionary .

马氏距离在模拟电路硬故障检测中的应用研究

Research on the application of Mahalanobis distance in catastrophic faults detection of analog circuits

计算机辅助编辑硬故障字典的优化问题

Optimized Computer-aided Compilation of Hard Failure Dictionary

诊断的试验结果表明，对于单、硬故障可以定位到元件级。

Experiments show that the location for the single and hard fault can be at element level .

介绍电脑硬盘软故障和硬故障的现象及处理方法。

This paper presents the phenomena of soft and hard faults of computer harddisk and their treatment methods .

提出了用故障字典法诊断模拟电路硬故障时，基本元件的物理故障和集成电路的功能故障进行故障建模的方法。

A fault modeling method of physical fault of basic element and functional fault of integrated circuit is proposed .

硬盘的故障从性质上分硬故障与软故障两大类。

The common fall of hard disk usually fall into ho kinds in property : damages to hard disk and software faults .

然后从实际芯片的缺陷类型中抽象出了丢失物缺陷、冗余物缺陷以及介质针孔缺陷的几何模型，给出了缺陷的粒径分布模型和由缺陷引起的软故障和硬故障的模型。

At last , the model of defect size distribution and the model of soft and hard faults caused by defects is given .

该方法不仅能够同时诊断元件的硬故障和参数偏移故障，而且能够同时运用于直流测试和交流测试。

The presented approach can not only diagnose both catastrophic faults and parametric faults , but also can be applied for both DC test and AC test .

软故障往往是硬故障的先兆，因此对制动系统软故障的及时预报和诊断是提升机故障诊断特点和难点。

Soft failures are often a precursor to hard failures , so the real-time forecasts of soft fault diagnosis are the characteristics and difficult of large complex systems .

该框架可以同时建立多个模拟印制板的诊断系统，可实现对直流状态下的硬故障诊断。

In this frame a number of printed analog circuit board troubleshooting systems can be established at the same time , which enables the detection of hardware fault under DC condition .

利用容差模拟电路节点电压灵敏度序列守恒定理，得到了模拟电路元件的软、硬故障统一样本。

In this paper , the unified sample for both soft and hard faults of elements in analog circuits is found according to the invariance of node-voltage sensitivity sequence in analog circuits with tolerance .

本文在区分缺陷与故障两个概念的基础上，将缺陷区分为成品率缺陷（硬故障）、可靠性缺陷（软故障）和良性缺陷。

Defects can be divided into yield defects ( hard faults ), reliability defects ( soft faults ) and goodness defects based on the difference between the defect 's concept and fault 's concept .

探讨了利用速率偏频激光捷联惯导／挠性捷联惯导陀螺的混合测量信息，通过神经网络实现硬故障交互检测的方法。

The problem about using the fused information of SINS based on rate-bias laser gyroscope / dynamical tune gyroscope to realize the component level hard fault interactions detection through a Back-Propagation neural network is discussed .

本文在传统的大规模电路的撕裂算法的基础上，提出了基于神经网络和交叉撕裂算法的大规模容差模拟电路模块级硬故障快速诊断方法。

This paper presents a fast approach of module level fault diagnosis for large-scale tolerance analog circuit based on the neural network and the crossover tearing technology according to conventional circuit decomposition technology in large-scale circuits .

仿真结果表明，该方法借助小波变换强大的时频分析能力，可以准确判定传感器软、硬故障，有效降低误报率和漏报率，具有良好的应用价值。

The simulation results prove that wavelet analysis has a superior skill of time and frequency analysis to detect sensor 's soft and hard failure , reduce distortion and failing report , and is effective and useful .

针对本文研究系统的特性，主要对传感器节点电源模块的硬故障和传感器采集数据异常的软故障进行诊断研究。

For the study of characteristics of the system , the fault diagnosis in research mainly includes the hardware fault caused by the power supply module for sensor nodes , and the soft fault caused by the unusual sensor data collected by sensor nodes .

基于模拟电路元件硬故障和软故障统一描述的思想，提出将电路测试节点电压信号的偏斜度与标准差组成的向量作为故障特征向量的方法。

Based on the idea of unified description of hard faults and soft faults of analog circuits , the vectors consisted of the skewness and standard deviations ( STDs ) of the test node voltages of analog circuits are proposed to form fault features .

在此理论基础上提出了节点电压灵敏度权序列故障字典法，该方法构造故障字典容易，故障字典规模小，能够诊断线性模拟电路软故障和硬故障。

According to these theorems fault dictionary method based on Node Voltage Sensitivity weight sequence is presented . By this method , it is easy to construct a fault dictionary of small size which can locate soft or hard fault in linear analog circuit .

实验结果表明，最优小波包变换适合于软故障特征提取，不完全小波包变换适合于硬故障特征提取。2、支持向量机与自适应共振理论相结合的多类分类算法的研究。

Experimental results show that the optimal wavelet packet transform suitable for soft fault feature extraction , incomplete wavelet packet transform suitable for hard fault feature extraction . 2 ^ Research on support vector machine combined with the adaptive resonance theory of multi-class classification algorithm .

它利用了模糊量的冗余性，克服了容差的影响，不仅能对硬故障，而且对软故障都能较好地定位到元件级，同时又可在线或离线诊断。

It uses the redundance of the fuzzy values , and overcomes the effect of the tolerance . The faults which are either hard faults or soft faults could be localized on the component rank , and the method can diagnosis whether on line or not .

文中将神经网络的自学习和分类技术应用于非线性电路直流故障诊断，把反向传播（BP）网络训练成一部能诊断软、硬单故障的故障字典。

In this paper , the self learning ability and the classification function of the artificial neural network have been used to nonlinear DC fault diagnosis , training a back propagation ( BP ) network as a fault dictionary that can diagnose single soft and hard faults .

但由于人为因素、计算机硬软件故障、病毒等都会造成数据的丢失，造成重大损失。

However , due to human factors , computer hardware disaster and software failures , virus , etc. will result in the loss of data , resulting in significant losses .

并以二次电源控制的容错设计为例，模拟电源控制器的各种硬软件故障，结果表明该方法能全面验证卫星在轨飞行时电源控制系统的容错能力和可靠性。

An example about fault-tolerant design of second power control system is given ; by simulate the software and hardware fault of power control system , it is proved that this method can completely verify the fault-tolerant capability and reliability of fault-tolerant design of satellite power control system in orbit .

我们研制的实验性的自动诊断系统，能准确地诊断出线性电阻网络中软、硬两类故障。

This experimental diagnostic system can correctly diagnose both soft and hard faults for linear resistive networks .