基于加速退化试验多性能参数相关性失效的可靠性评估模型

考虑加速退化试验中多性能参数之间的相关性失效,对产品在正常工作应力的可靠性进行评估.结合数理统计和Copula函数相关理论,提出了一种基于加速退化试验多关键性能参数相关性失效产品可靠性评估模型.对多元加速退化试验数据进行统计分析可以获得高应力下的多性能参数退化规律,再外推至正常水平,对产品的可靠性进行评估.     

基于比例风险退化模型的产品可靠性评估

为了利用退化数据评估长寿命产品的可靠性,提出一种新的比例风险退化模型.利用时间退化数据与应力加速寿命之间的相似关系,并假定退化数据的失效概率函数是时间协变量的比例函数,从而实现基于比例风险退化模型的可靠性评估.采用最小二乘法估计模型参数.由已知时刻的退化量失效率及比例函数外推产品在特定时间、预定失效阈值下的失效率及可靠度函数.以一种长寿命GaAs激光器的退化数据为例,完成该产品的可靠性评估.结果表明：所提方法能准确描述退化过程,是长寿命产品可靠性评估的一种有效方法.     

考虑相关竞争故障过程及变动阈值的可靠性评估(英文)

研究了受退化和随机冲击共同作用的产品可靠性评估问题.考虑相关竞争故障过程,建立了2类概率模型.首先,基于累积冲击假设,分别建立了硬故障概率模型和软故障概率模型.以此为基础,考虑退化和冲击过程建立了相关竞争故障模型.此外,研究了因多次冲击载荷作用而导致的硬故障阈值下降的情形,分析故障阈值的变动对产品可靠性的影响.以一组裂纹增长数据为例,验证模型的有效性.研究结果表明:冲击对产品故障过程有着显著影响,同时硬故障阈值的下降会进一步加剧上述效应.     

退化失效线性模型及统计分析

利用退化试验进行高可靠长寿命的产品可靠性分析是一条行之有效的途径.基于产品性能退化量,根据绝对和相对失效标准下的产品寿命,分别讨论了退化失效下的随机截距和随机斜率线性模型,并给出了随机参数服从正态和Weibull分布情形下的结果.     

基于PCA-CMAC的设备性能退化评估

提出了一种用于设备性能退化评估的PCA-CMAC(主成分分析-小脑模型节点控制器)模型.该模型利用PCA进行特征提取,去除多个传感器信号特征的冗余信息,并且减少CMAC的输入维数;利用CMAC的局部泛化能力定量地评估设备的性能退化.给出了模型的实现过程,并将模型应用于钻削过程刀具状态的评估,试验结果证明该模型能基于刀具的正常状态,对刀具的磨损状态进行定量的评估.分析了CMAC中泛化参数g和量化参数r对评估结果的影响,g越大,CMAC的泛化能力越好,但各退化状态之间的区别越不明显;r越小,各退化状态之间越容易区分,但所需的权存储空间越大.2个参数的基本选择原则是CMAC的权存储空间应尽量小,与此同时,各退化状态之间应容易区分.     

基于威布尔过程的长寿命产品退化数据统计分析

基于传统的伪寿命分布、退化量分布、随机系数分布这三种方法进行的高长寿命产品可靠性评估,往往会产生较大偏差.为了较精确的估计出高长寿命产品的寿命可靠度,通过分析产品的退化失效机理,利用威布尔过程拟合产品相邻单位退化量的时间间隔变化趋势,可以由产品的试验数据外推出产品的寿命,进而可以较精确地预计产品的剩余寿命等可靠性指标.     

大跨度斜拉桥拉索安全评估的概率方法

为了对运营期斜拉桥安全性进行评估,提出了评估斜拉索安全性的方法.考虑Daniels效应和断丝概率提出斜拉索强度模型与基于斜拉索运营期安全系数的大跨度斜拉桥可靠度评估方法.作为模型的应用验证,分析了宁波招宝山斜拉桥最长边索,即第25号索在不同检测期的安全性,并调查了不同的参数对安全概率的影响.分析结果表明:斜拉索Daniels效应不可忽略;运营期斜拉索的退化可以通过线性增长的索内断丝率来模拟;考虑斜拉桥拉索退化过程的随机性评估运营期斜拉桥拉索安全水平的方法是有效的.     

基于VMD和嵌入选择NPE的滚动轴承性能退化评估（英文）

为了提高滚动轴承性能退化评估中退化指标的早期故障敏感性和稳定性,提出了一种基于嵌入选择的邻域保持嵌入(ESNPE)方法.首先,采用变分模态分解(VMD)对获得的振动信号进行分解,提取各本征模态分量的奇异值和相对能量等组成高维故障特征集.然后,采用NPE流行学习方法提取特征空间内的嵌入特征.针对传统NPE存在有效嵌入信息容易被抑制的问题,构建了一种基于Spearman相关系数的嵌入选择策略.该策略通过相关系数的大小衡量嵌入特征的有效性,并通过一阶差分的方法在轴承退化的早期阶段确定并保留有效嵌入特征.最后,采用支持向量数据描述(SVDD)模型构建性能退化指标,实现轴承性能退化评估.使用轴承全寿命退化实验数据,并与传统的主成分分析(PCA)方法和NPE方法特征提取分析结果进行对比,验证了所提方法在提升退化指标早期故障敏感性和稳定性方面具有优越性.     

基于时间序列的退化型产品可靠性评估方法

在不假定退化轨道的基础上,提出利用时间序列的方法,使用ARIMA模型对退化数据进行建模,拟合其退化规律,预测失效寿命,进而进行可靠性评估.并给出实例,说明该方法具有较好的适用性和有效性.     

基于有限元仿真和可靠性理论的腐蚀管道剩余寿命评估

提出一种腐蚀石油管道剩余使用寿命的综合评价方法.考虑泄漏和爆裂2种失效模式,利用结构可靠性理论建立相应的极限状态函数.采用幂律函数建立腐蚀缺陷增长模型,分析外部环境因素对管道缺陷增长的影响,并将计算结果与常用的线性增长模型分析结果进行比较.建立含腐蚀缺陷管道爆破压力的有限元仿真模型,通过与国际通用标准和水压爆破试验结果...     

基于动态可靠度的在役桥梁维修决策经济优化分析

The lifeti me maintenance of a bridge can comprise amuch greater portion of the total lifeti me cost than theoriginal cost of construction because of deterioration.An opti mal maintenance programis the key to makingappropriate decisions to mini mize cost and maintain anappropriate level of safety.Life-cycle cost is a usefulconcept in reducing the overall cost and achieving anappropriate maintenance plan[1,2].It is difficult tomake a reliable esti mate of thelifeti me costs whenthefuture of the…     

静载荷受剪螺栓连接的可靠性设计及分析

分析静载荷受剪螺栓连接的主要失效模式,着重论述了其可靠性设计过程,并与传统设计方法进行了比较分析,从而阐明了可靠性设计的特点和优势.     

Nonparametric modeling on random uncertainty and reliability analysis of a dual-span rotor

A general procedure is proposed to estimate the reliability of a dual-span rotor based on nonparametric modeling on random uncertainty. First, the vibration equation of the rotor with random uncertainty is constructed based on random matrices through the nonparametric modeling approach. Second, the reliability estimation is then performed by response spectral analysis and the moment method. By making full use of the advantages of nonparametric method and response spectral analysis, not only is the requirement on probability density function (PDF) avoided, but also the first and second moments are no longer needed to be estimated or assumed for calculating the reliability. Finally, the statistical index Z*-value based on short-term predictability is introduced to investigate the influence of random uncertainties on the reliability of the dual-span rotor. Illustrating examples show that the results obtained from the proposed procedure are consistent with those from short-term predictability, such that dangerous ranges can be well identified during the start-up process of the rotor.     

Quantification for complex assessment: uncertainty estimation in final year project thesis assessment

A quantitative method for estimating an expected uncertainty (reliability and validity) in assessment results arising from the relativity between four variables, viz examiner's expertise, examinee's expertise achieved, assessment task difficulty and examinee's performance, was developed for the complex assessment applicable to final year project thesis assessment including peer assessment. A guide map can be generated by the method for finding expected uncertainties prior to the assessment implementation with a given set of variables. It employs a scale for visualisation of expertise levels, derivation of which is based on quantified clarities of mental images for levels of the examiner's expertise and the examinee's expertise achieved. To identify the relevant expertise areas that depend on the complexity in assessment format, a graphical continuum model was developed. The continuum model consists of assessment task, assessment standards and criterion for the transition towards the complex assessment owing to the relativity between implicitness and explicitness and is capable of identifying areas of expertise required for scale development.     

Reduced Expanding Load Method for Simulation-Based Structural System Reliability Analysis

The current situation and difficulties of the structural system reliability analysis are mentioned. Then on the basis of Monte Carlo method and computer simulation, a new analysis method reduced expanding load method (RELM) is presented, which can be used to solve structural reliability problems effectively and conveniently. In this method, the uncertainties of loads, structural material properties and dimensions can be fully considered. If the statistic parameters of stochastic variables are known, by using this method, the probability of failure can be estimated rather accurately. In contrast with traditional approaebes,RELM method gives a much better understanding of structural failure frequency and its reliability index β is more meaningful. To illustrate this new idea, a specific example is given.     

电力二次设备继电保护系统失效模型研究

继电保护设备是一种重要的电力二次设备,继电保护装置分为继电保护装置本体和二次回路．文章通过对继电保护装置本体应用软／硬件失效模型进行风险评估,对二次回路应用故障树模型进行分析,建立了继电保护系统失效模型,该模型能够帮助电力自动化系统较准确地判断出系统的故障率和运行可靠性．     

Nonlinear Dynamic Reliability of Coupled Stay Cables and Bridge Tower

Nonlinear vibration can cause serious problems in long span cable-stayed bridges.When the internal resonance threshold is reached between the excitation frequency and natural frequency,large amplitudes occur in the cable.Based on the current situation of lacking corresponding constraint criteria,a model was presented for analyzing the dynamic reliability of coupling oscillation between the cable and tower in a cable-stayed bridge.First of all,in the case of cable sag,the d'Alembert principle is applied to studying the nonlinear dynamic behavior of the structure,and resonance failure interval of parametric oscillation is calculated accordingly.Then the dynamic reliability model is set up using the JC method.An application of this model has been developed for the preliminary design of one cable-stayed bridge located on Hai River in Tianjin,and time histories analysis as well as reliability indexes have been obtained.When frequency ratio between the cable and tower is approaching 1∶2,the reliability index is 0.98,indicating high failure probability.And this is consistent with theoretical derivation and experimental results in reference.This model,which is capable of computing the reliability index of resonance failure,provides theoretical basis for the establishment of corresponding rule.