竞争硬故障与软故障失效模式产品的可靠性分析

根据系统可靠性理论,考虑到硬故障和软故障两种竞争失效模式对产品失效的影响,对传统的可靠性分析方法进行了扩展。研究了满足硬故障和软故障竞争失效规则的产品可靠性函数和失效率函数。并通过不同的失效模式之间的关系。得到产品在不同的硬、软故障模式下的失效分布函数。另外还给出一个简单的算例来说明该分析方法的应用。该分析方法可能更符合机电产品的失效规律。     

电磁脉冲应力下MOS器件退化失效建模

通过分析电磁脉冲应力下,栅氧化层软击穿(MOS器件主要失效模式)的失效机理,得出结论:它符合基于随机过程的退化失效模型。根据此结论,提出利用该模型来描述电磁脉冲应力下MOS器件的退化失效过程,并给出相应的退化失效模型。同时针对退化失效模型中的失效阈值问题,研究了随机失效阈值问题,分析了周期电磁脉冲应力下MOS器件的失效阈值问题,给出动态应力-强度干涉(SS)I模型。这些为更合理描述和分析MOS器件的退化失效问题提供了新的途径。     

Equivalence of Redundant Systems With Respect to Time to Failure

The time to failure of an m out of n component system, with component failure times exponentially and identically distributed, is treated. This failure time is shown to be identically distributed to the failure time of an n - m + 1 component standby system, with component failure times exponentially but not identically distributed. The proof of this result is based on a recursion formula for the distribution of time to system failure.     

基于半马尔科夫过程的变压器故障率预测及研究

设备的故障率曲线是制定维修策略的重要指标之一。为了合理估计变压器的故障率,本文提出了基于半马尔科夫过程的预测模型。模型中考虑了变压器的热老化过程和随机故障过程。热老化过程根据绝缘纸平均聚合度的取值范围划分为4个状态,状态间的转移时间假设服从威布尔分布,威布尔分布的尺度参数由经验公式获得;随机故障过程考虑了由雷击和短路冲击引起的2个随机故障状态,老化状态和随机故障状态之间的转移时间假设服从指数分布,转移率由故障统计数据获得。对模型中威布尔分布的形状参数进行适当取值后,可获得符合实际统计结果的变压器故障率曲线。     

A prognostic method for predicting failure of dc/dc converter

The failure of dc/dc converters can directly result in electronic systems working unconventionally or significant downtime. To pre-determine time to failure and generate substantial safety and cost benefits, it is necessary to assess the extent of deviation of dc/dc converters from its expected state of health in real time and predict time to failure in advance. This paper presents a novel prognostic method for predicting the time to failure of dc/dc converters. The process involves identifying precursor parameters, determining prognostic of failure, and determining a criterion for predicting time to failure. The output voltage is used as a precursor parameter and directly monitored when the converter with a given load periodically operates at different temperature stresses. The phenomenon that the differences of output voltages collected at different temperature stresses begin to increase with a large (or small) fluctuation is detected in collected output voltages. This phenomenon is identified as a prognostic of failure. A percentage of the initial difference is used as the criterion for predicting time to failure. A case study is given to illustrate the procedure that how to monitor output voltages, detect prognostic and predict time to failure. The results show the health state could be assessed in real time and the time to failure could be predicted in advance. Furthermore, the deviation of the predicted time to failure from the actual time to failure could meet the demand of a considered acceptable range in engineering practice.     

Comparing cumulative incidence functions of a competing-risks model

A competing-risks model refers to a situation where a system (or organism) is exposed to two or more causes of failure (or death) but its eventual failure (or death) can be attributed to exactly one of the causes of failure. The basic information available in the competing-risks situation is the time to failure of the system, and the corresponding cause of failure. In practice, the causes of failure are often statistically dependent (the latent failure time of an individual failing from one cause of failure is statistically correlated with the latent failure time of the same individual failing from a different cause of failure). This paper provides a simple nonparametric hypothesis test (SNPHT) for comparing the cumulative incidence functions of a competing-risks model when two causes of failure are possibly statistically dependent. The test statistic is the weighted sum of the differences of two cumulative incidence functions at system failure times. This paper: (1) proves that the test statistic has asymptotic normal distributions under both null and alternative hypotheses; and (2) derives an explicit formula for the power function of SNPHT. The simulation study for SNPHT, based on the absolutely continuous bivariate exponential model, shows that the simulated powers and the approximated powers calculated from the formula are consistent for a moderate sample size. SNPHT is very easy to use. The illustrative example involves the failure of small electrical appliances     

拜占庭容错服务的适应性失效检测研究

文章在Byzantine fault tolerant状态机复制容错算法（BFT）中增加适应性失效检测。根据拜占庭容错服务应用需求和运行环境动态调整失效检测时间，适应性好，能满足BIT失效检测要求。当主复件失效时,可以缩短视图切换时间，减少系统响应时间，提高系统可用性。     

基于缺陷数据的雷达软件失效模式分析与应用

基于雷达软件缺陷历史数据开展失效模式分析与应用研究。首先,结合雷达软件特点与失效机理,构建失效数据模型;然后,在失效数据模型基础上,阐述雷达软件失效模式分析与应用方法,以及相应的算法原理;最后,开展工程实例研究,获取适用于雷达软件的典型失效模式。实例结果表明:文中所提出的方法可有效提升雷达软件失效分析的效率和质量,避免相似问题重复发生,保障雷达软件质量满足系统要求。     

Fault-Tree Analysis by Fuzzy Probability

In conventional fault-tree analysis, the failure probabilities of components of a system are treated as exact values in estimating the failure probability of the top event. For many systems, it is often difficult to evaluate the failure probabilities of components from past occurrences because the environments of the systems change. Furthermore, it might be necessary to consider possible failure of components even if they have never failed before. We, therefore, propose to employ the possibility of failure, viz. a fuzzy set defined in probability space. The notion of the possibility of failure is more predictive than that of the probability of failure; the latter is a limiting case of the former. In the present approach based on a fuzzy fault-tree model, the maximum possibility of system failure is determined from the possibility of failure of each component within the system according to the extension principle. In calculating the possibility of system failure, some approximation is made for simplicity.     

On optimal burn-in procedures - a generalized model

Burn-in is a manufacturing technique that is intended to eliminate early failures. In this paper, burn-in procedures for a general failure model are considered. There are two types of failure in the general failure model. One is Type I failure (minor failure), which can be removed by a minimal repair or a complete repair; and the other is Type II failure (catastrophic failure), which can be removed only by a complete repair. During the burn-in process, two types of burn-in procedures are considered. In Burn-In Procedure I, the failed component is repaired completely regardless of the type of failure; whereas, in Burn-In Procedure II, only minimal repair is done for the Type I failure, and a complete repair is performed for the Type II failure. Under the model, various additive cost functions are considered. It is assumed that the component before undergoing the burn-in process has a bathtub-shaped failure rate function with the first change point t/sub 1/, and the second change point t/sub 2/. The two burn-in procedures are compared in cases when both the procedures are applicable. It is shown that the optimal burn-in time b/sup */ minimizing the cost function is always before t/sub 1/. It is also shown that a large initial failure rate justifies burn-in, i.e., b/sup */>0. The obtained results are applied to some examples.     

雷达故障检测专家系统设计

新体制雷达系统的复杂性增加了雷达故障检测的难度。为保证雷达工作的可靠性和维修性能,首先根据雷达设备故障检测的特点,运用人工智能理论,提出了基于神经网络和专家系统相结合的检测技术,并把这种技术应用到雷达的故障检测中。概述了系统的基本结构,介绍了系统的功能、知识的获取和存储、推理机制、解释机制等,并给出了具体的检测实例,阐述了雷达故障检测专家系统组成和基于神经网络专家系统雷达发射系统故障检测。结果证明,利用神经网络技术建立的故障检测专家系统可提高雷达故障检测的正确性和效率。     

Stochastic analysis of a fault tolerant network system

This paper deals with a stochastic model related to a converter-based communication network system. In this model two converters are connected in parallel with a network system. Each converter has two failure modes—transient failure and latent failure. Transient failures occur due to minor faults in converters. Further the converter may enter into the latent failure mode if major hardware problems arise in the transient failure mode. The converter may recover automatically when it is in a transient failure mode and needs repair when it is in a latent failure mode. The system also enters preventive maintenance at random epochs when both converters are normal. System failure occurs when both the converters are in a latent failure mode. Using the regenerative point technique, various reliability measures are obtained.     

Fractional duration before first failure—A useful index and an analytical tool

This paper defines a new index, fractional duration before first failure and describes the methodology for its calculation. The relationship of this index to the mean time to first failure is also presented. The fractional duration before first failure is, therefore, not only a useful measure of system reliability but also provides an alternative approach to calculate mean time to first failure. The concepts and techniques are illustrated with suitable examples.     

基于失效物理的集成电路故障定位方法

超大规模集成电路后道工艺(BEOL)中的失效日益增多,例如多层金属化布线桥连、划伤,栅氧化层的静电放电(ESD)损伤、裂纹等失效模式,由于失效点本身尺寸小加上电路规模大,使得失效分析难度增加。为了能够对故障点进行快速、精确定位,提出了基于失效物理的集成电路故障定位方法。根据CMOS反相器电路的失效模式提出了4种主要故障模型:栅极电平连接至电源(地)、栅极连接的金属化高阻或者开路、氧化层漏电和pn结漏电。结合故障模型产生的光发射显微镜(PEM)和光致电阻变化(OBIRCH)现象的特征形貌和位置特点,进行合理的失效物理假设。结果表明,基于该方法可对通孔缺陷、多层金属化布线损伤以及栅氧化层静电放电损伤失效进行有效的定位,快速缩小失效范围,提高失效分析的成功率。     

New application of laser beam to failure analysis of LSI with multi-metal layers

A new technique for failure analysis of LSI with multi metal layers is described. There is a fabrication technique to form a big and optional window on upper layers using Nd-YAG laser without destroying any electrical function, and to approach the failure point through this window. The failure analysis procedure based on logical flow is presented. Fabrication technique is located in part of this procedure which consists of four steps. Two difficult reasons for approaching the failure point without fabrication technique are described. This difficulty results from line composition and line width. The fabrication procedure using Nd-YAG laser is reported. This procedure is to cover the chip surface with photo resist, and form a big and optional size window on upper layer with laser beam, and finally expose the purposed layer. Three failure analysis examples using this technique are introduced: unformed contact falure mode, Si-noduled failure mode, and isolation destroying failure mode.     

适应性失效检测器研究与设计

研究网络延时和丢包对失效检测器服务质量的性能影响,设计了一种更符合网络环境变化的适应性失效检测器。新的适应性失效检测器不仅能更好地适应网络延时情况,也能适应不同应用系统对网络丢包情况的要求。测试结果表明,该适应性失效检测器能有效降低错误率,进一步提高失效检测服务质量。     

基于SIFT技术的集成电路失效缺陷分析

激励源诱导故障测试(SIFT)是一种新型的失效定位技术,可用于集成电路和分立器件中漏电、击穿、短路等失效点的定位及失效机理的分析。在介绍SIFT技术工作原理的基础上,利用该技术进行了六反相器电路的深埋层缺陷、收发器电路中电源与地之间漏电流失效和串行输出模数转换电路MOS器件欧姆短路的定位,并结合微结构观测分析了失效原因。研究结果表明,SIFT技术能有效分析光发射显微镜(EMMI)和激光光束诱导阻抗变化测试(OBIRCH)技术较难定位的缺陷,弥补了这些常规失效分析技术的不足。     

Consecutive k-out-of-n:F system with sequential failures and a single repair

This paper presents a solution for determining the mean time to system failure of a consecutive k-out-of-n:F system with a single repair. The components of the system are subject to sequential failures. The mean time to system failure is evaluated by making use of the relation between the reliability function and the probability of the first passage time to system failure. The Laplace transform of the first passage time to system failure is obtained as a ratio of two determinants.     

基于MEMS工艺的微机械失效分析研究进展

随着微电子机械系统(MEMS)消费市场的不断增长,微器件的可靠性问题将是未来数年MEMS研究面临的新挑战.针对国内外近年来在微机械失效方面的主要研究内容,综述了基于MEMS工艺微机械的主要失效形式、失效机理和失效预防措施的最新研究成果和主要研究方法,着重总结了微机械的断裂、疲劳、磨损、黏附和蠕变等失效形式,并分析了这些失效形式对微器件的功能和性能的影响以及对这些失效的预防措施,通过结构优化、器件设计和加工质量控制等措施可降低失效的发生,提高可靠性.最后总结了微机械失效分析中存在的主要问题和以后的发展方向,提出建立通用的失效预测模型、标准的可靠性测试方法是未来微机械失效分析的研究重点.     

Insufficiency of Instantaneous Strength Determinations for Failure Rate Prediction

The failure of an element subjected to a steady applied stress is associated with the time-wise deterioration of the element's strength. At the instant of failure, the element can no longer perform to specifications. The strength of the element can be obtained at a specified point in time by abruptly modifying the applied stress to a level which causes the element to fail. The instantaneous strengths determined by this procedure are investigated for the possibility of reducing the test time necessary in obtaining the failure rate associated with exponentially distributed failure times. A simple, physically plausible class of strength deterioration functions is considered and the distributional character of the instantaneous strengths at any time instant is determined. It is found that when the underlying distribution of time to failure is exponential, the instantaneous strengths have distributions which are invariant with time. This result indicates that no clue to the form of strength deterioration can be obtained by testing for the instantaneous strengths. Thus the conclusion is reached, in the case of the exponential failure distribution, that instantaneous destruction tests for strength determination are ineffective for assessing the underlying failure rate.