Semiparametric inference for the accelerated life model with time-dependent covariates

1995 ◽  
Vol 44 (1) ◽  
pp. 47-63 ◽  
Author(s):  
D.Y. Lin ◽  
Zhiliang Ying
Keyword(s):  
Time Dependent ◽  
Semiparametric Inference ◽  
Accelerated Life ◽  
Life Model ◽  
Time Dependent Covariates ◽  
Accelerated Life Model

scholarly journals Design and application of accelerated life test of grating ruler based on step stress loading

2021 ◽  
Vol 336 ◽  
pp. 02027
Author(s):  
Jili Wang ◽  
Qingyu Li ◽  
Xiaocui Zhu ◽  
Cheng Gao ◽  
Yi Li ◽  
...  
Keyword(s):  
Weibull Distribution ◽  
Test Method ◽  
Accelerated Life Test ◽  
Test Scheme ◽  
Life Test ◽  
Test Device ◽  
Typical Application ◽  
Accelerated Life ◽  
Life Model ◽  
Accelerated Life Model

Combined with the actual project, a grating ruler accelerated life test device is designed, which can simulate the actual loads, including temperature, humidity and speed stress. An accelerated life test scheme based on stepped stress loading is proposed, and 6 grating rulers are tested based on time-censored test method. An accelerated life model based on Weibull distribution used for evaluating the lifetime is established. Related experimental techniques also can be typical application cases for innovative practical teaching.

Composite GM (1, 1) Model for Treating Data of Constant Stress Amplitude Accelerated Life Tests

2008 ◽  
Vol 44-46 ◽  
pp. 859-862
Author(s):  
Yi Qian ◽  
Y.H. Zhou ◽  
J.C. Pei
Keyword(s):  
Stress Amplitude ◽  
Constant Stress ◽  
Mixed Data ◽  
Grey Theory ◽  
Accelerated Life Tests ◽  
Accelerated Life ◽  
Life Model ◽  
Life Tests ◽  
Accelerated Life Model ◽  
Model Approach

An accelerated life model is an important factor affecting the evaluation of test data. Avoiding the existent difficulties of treating mixed data and based on the grey theory with composite GM(1,1) model, approach is presented for treating fatigue data under constant stress amplitude accelerated life tests. Availability of the approach has indicated by an example.

BINARY LOGISTIC REGRESSION AND PHM ANALYSIS FOR RELIABILITY DATA

2014 ◽  
Vol 21 (05) ◽  
pp. 1450023 ◽  
Author(s):  
ALEXANDRE C. MENDES ◽  
NASSER FARD
Keyword(s):  
Logistic Regression ◽  
Survival Data ◽  
Recurrent Events ◽  
Binary Logistic Regression ◽  
Time Dependent ◽  
Time Interval ◽  
Data Set ◽  
Accelerated Life ◽  
Address Data ◽  
Time Dependent Covariates

This study proposes a modification for the binary logistic regression to treat time-dependent covariates for reliability studies. The proportional hazard model (PHM) properties are well suited for modeling survival data when there are categorical predictors; as it compares hazards to a reference category. However, time-dependent covariates present a challenge for the analysis as stratification does not produce hazards for the covariate stratified or creation of dummy time-dependent covariates faces difficulty on selecting the time interval for the interaction and the coefficient results may be difficult to interpret. The findings show that the logistic regression can provide equal or better results than the PHM applied for reliability analysis when time-dependent covariate is evaluated. The PHM is potentially preferred to address data set without time-dependent variables as it does not require any data manipulation. The logistic regression ignores the information on timing of the events; which is corrected by breaking each subject survival history into a set of discrete time intervals that are treated as distinct observations evaluated as a binary distribution. Recurrent events can be addressed by both methods with proper correction for lack of heterogeneity. The application of the modified logistic regression model for the study of reliability is innovative and with readily potential application for step-stress time-dependent accelerated life testing.

Sign in / Sign up

Export Citation Format

Share Document