ABSTRACT

This chapter discusses the problem of inferences about the parameters of the process and the issues related to prediction that arise when the Power law process (PLP) is used to model reliability growth. It discusses various methods of confidence interval estimation for the parameters. The chapter presents the confidence intervals based on future observations or predictive distribution. It also presents a detailed trend test and tests for parameter-based inferences. The chapter also discusses the issues related to reliability estimation. PLP or Weibull process is quite often used in developing repairable system models. The conditional inference on the PLP model was first explored by K. Muralidharan et al., where they have provided interval estimation for the ratio of intensity parameters. The adequacy of the PLP for a particular data set can be diagnosed graphically using either Duane plots or modified total time on tests plots.