An Extension to the Weibull Process Model
ARMY MATERIEL SYSTEMS ANALYSIS ACTIVITY ABERDEEN PROVING GROUND MD
Pagination or Media Count:
A system testing program may consist of two kinds of test phases those in which system configuration is continually modified to reduce the frequency of failures and those in which system configuration does not change. The system failure rate may be modeled as the intensity function of a Weibull process in the former case and a homogeneous Poisson process in the latter case. This paper considers a nonhomogeneous Poisson process consisting of a time truncated Weibull process followed by a time truncated homogeneous Poisson process and illustrates how such an extended model can be used to plan and evaluate a reliability demonstration test. The resulting methodology allows a unified treatment of failure data from two consecutive test phases and can reduce consumer and producer risks while at the same time requiring less test time.
- Statistics and Probability
- Manufacturing and Industrial Engineering and Control of Production Systems