JMP 13.2 Online Documentation (English)
Discovering JMP
Using JMP
Basic Analysis
Essential Graphing
Profilers
Design of Experiments Guide
Fitting Linear Models
Predictive and Specialized Modeling
Multivariate Methods
Quality and Process Methods
Reliability and Survival Methods
Consumer Research
Scripting Guide
JSL Syntax Reference
JMP iPad Help
JMP Interactive HTML
Capabilities Index
JMP 12 Online Documentation
Reliability and Survival Methods
•
Reliability Growth
• Reliability Growth Platform Overview
Previous
•
Next
Reliability Growth Platform Overview
The Reliability Growth platform fits Crow-AMSAA models, described in MIL-HDBK-189 (1981). A Crow-AMSAA model is a non-homogeneous Poisson process (NHPP) model with Weibull intensity; it is also known as a power law process. Such a model allows the failure intensity to vary over time. The failure intensity is defined by the two parameters beta and lambda.
For single-prototype data, the platform fits four classes of models and performs automatic change-point detection. The following reports are available:
•
Simple Crow-AMSAA model, where both parameters are estimated using maximum likelihood.
•
Crow-AMSAA with Modified MLE, where the maximum likelihood estimate for beta is corrected for bias.
•
Fixed Parameter Crow-AMSAA model, where the user is allowed to fix either or both parameters.
•
Piecewise Weibull NHPP model, where the parameters are estimated for each test phase, taking failure history from previous phases into account.
•
Reinitialized Weibull NHPP model, where both parameters are estimated for each test phase in a manner that ignores failure history from previous phases.
•
Automatic estimation of a change-point and the associated piecewise Weibull NHPP model, for reliability growth situations where different failure intensities can define two distinct test phases.
For multiple-prototype data, the platform fits the following classes of models:
•
Piecewise Weibull NHPP model, where each system in a multi-phase study follows the same piecewise Weibull NHPP model. The differences among the systems are assumed to be due to the randomness of individual realizations of the same model. This model contains one beta parameter for each phase and one lambda parameter.
•
Piecewise Weibull NHPP with Different Intercepts model, where each system in a multi-phase study follows a separate piecewise Weibull NHPP model. This model contains one beta parameter for each phase and one lambda parameter for each system.
•
Distinct Phase Weibull NHPP model, where each system in a multi-phase study follows the same Crow-AMSAA model in each phase. This model contains one beta parameter and one lambda parameter for each phase.
•
Distinct Weibull NHPP model, where each system in a multi-phase study follows a separate Crow-AMSAA model in each phase. This model contains one beta parameter and one lambda parameter for each combination of system and phase in the study.
•
Distinct System Weibull NHPP model, where each system in the study follows a separate Crow-AMSAA model with different parameters.
•
Identical System Weibull NHPP model, where each system in the study follows a single Crow-AMSAA model. The differences among the systems are assumed to be due to the randomness of individual realizations of the same model.
Interactive profilers enable you to explore changes in MTBF, failure intensity, and cumulative failures over time. When you suspect a change in intensity over the testing period, you can use the change-point detection option to estimate a change-point and its corresponding model.
Previous
•
Next
Help created on 9/19/2017