International Journal of Reliability, Quality and Safety Engineering

Reliable software is mandatory for complex mission-critical systems. Classifying modules as fault-prone, or not, is a valuable technique for guiding development processes, so that resources can be focused on those parts of a system that are most likely to have faults.

Logistic regression offers advantages over other classification modeling techniques, such as interpretable coefficients. There are few prior applications of logistic regression to software quality models in the literature, and none that we know of account for prior probabilities and costs of misclassification. A contribution of this paper is the application of prior probabilities and costs of misclassification to a logistic regression-based classification rule for a software quality model.

This paper also contributes an integrated method for using logistic regression in software quality modeling, including examples of how to interpret coefficients, how to use prior probabilities, and how to use costs of misclassifications. A case study of a major subsystem of a military, real-time system illustrates the techniques.

The investigation of approximate expressions for availability of complex repairable structures is an important problem facing researchers in the field of reliability engineering. This is due to the fact that the availability evaluation of complex structures becomes time consuming even under the assumption that the failure and repair rates are constant and only the steady-state availability and mean time to failure (MTTF) are needed. The main reason for complexity is due to the large number of states required to describe the system. In such cases, approximate expressions can be useful. In this paper, approximate expressions for steady-state availability and MTTF of large series–parallel structures are derived under three different assumptions on maintenance strategy, namely, independent elements, macro-structures and independent macro-structures. An example is presented to illustrate the application of the method under the three assumptions. Numerical evaluation of the example problem shows a good convergence of the results.

This paper considers the problem of selecting an optimal setting of the process mean and an optimal production run for a continuous production process. The process is subject to gradual shifts in the process mean due to occurrences of some random shocks. The product output becomes nonconforming only when the process experiences a certain number of accumulated shocks. The changes in the process mean are assumed to follow a nonhomogeneous Poisson process. A quadratic loss function, which is a general form of Taguchi's loss function, is utilized for developing the economic model in determining an initial resetting process mean and an optimal production run. Some new results are derived and some interesting findings are reported.

We present a method for selecting the most reliable population under step-stress accelerated life testing with type II censoring. We construct a new statistic, the transitional order statistic (TOS), and derive an approximate expression for its distribution. Using the TOS, a selection rule is formulated from the test results. For planning purposes, we establish the relation between sample size and the probability of correct selection by defining a nonlinear indifference zone under the least favorable configuration. Finally, a simulation study is performed to illustrate the selection procedure and to validate the associated probability of correct selection.

In this paper, a useful family of discontinuous point processes is proposed, which is able to analyze the failure pattern of repairable units subjected to repair actions that depart from the commonly assumed minimal repair policy. Maximum likelihood estimators of the parameters which index two special cases of the above family are derived, and procedures for testing the departure from the minimal repair assumption, based on asymptotic results, are discussed. An exact and unbiased testing procedure to be performed for small or moderate sample sizes is also proposed. Finally, numerical examples are given to illustrate the proposed estimation and testing procedures.

Greenwich and Jahr-Schaffrath⁹ introduced the incapability index C_pp, which is a simple transformation of the index proposed by Chan et al.³ Chen¹⁰ considered the incapability index , a generalization of C_pp, to handle processes with asymmetric tolerances. Based on the same idea on , we consider a new generalization , which is a modification of the process capability index C_pm. In the cases of symmetric tolerances, the new generalization reduces to the original index C_pm. The new generalization not only takes the proximity of the target value into consideration, like those of C_pm and , but also takes into account the asymmetry of the specification limits. We compare the new generalization with C_pa(1, 3) and C_pa(0, 4), two special cases of C_pa(u, v) recommended by Vännman⁷ for asymmetric tolerances. We also investigate the statistical properties of the natural estimator , assuming the process is normally distributed. We obtain the exact distribution and an explicit form of the probability density function of . In addition, we compute the rth moment-expected value, variance of , and analyze the bias as well as the MSE of .