Tadashi Dohi, Hiroyuki Okamura and Cun Hua Qian
In this paper, the authors propose two construction methods to estimate confidence intervals of the time-based optimal software rejuvenation policy and its associated maximum…
Abstract
Purpose
In this paper, the authors propose two construction methods to estimate confidence intervals of the time-based optimal software rejuvenation policy and its associated maximum system availability via a parametric bootstrap method. Through simulation experiments the authors investigate their asymptotic behaviors and statistical properties.
Design/methodology/approach
The present paper is the first challenge to derive the confidence intervals of the optimal software rejuvenation schedule, which maximizes the system availability in the sense of long run. In other words, the authors concern the statistical software fault management by employing an idea of process control in quality engineering and a parametric bootstrap.
Findings
As a remarkably different point from the existing work, the authors carefully take account of a special case where the two-sided confidence interval of the optimal software rejuvenation time does not exist due to that fact that the estimator distribution of the optimal software rejuvenation time is defective. Here the authors propose two useful construction methods of the two-sided confidence interval: conditional confidence interval and heuristic confidence interval.
Research limitations/implications
Although the authors applied a simulation-based bootstrap confidence method in this paper, another re-sampling-based approach can be also applied to the same problem. In addition, the authors just focused on a parametric bootstrap, but a non-parametric bootstrap method can be also applied to the confidence interval estimation of the optimal software rejuvenation time interval, when the complete knowledge on the distribution form is not available.
Practical implications
The statistical software fault management techniques proposed in this paper are useful to control the system availability of operational software systems, by means of the control chart.
Social implications
Through the online monitoring in operational software systems, it would be possible to estimate the optimal software rejuvenation time and its associated system availability, without applying any approximation. By implementing this function on application programming interface (API), it is possible to realize the low-cost fault-tolerance for software systems with aging.
Originality/value
In the past literature, almost all authors employed parametric and non-parametric inference techniques to estimate the optimal software rejuvenation time but just focused on the point estimation. This may often lead to the miss-judgment based on over-estimation or under-estimation under uncertainty. The authors overcome the problem by introducing the two-sided confidence interval approach.
Details
Keywords
Cun Hua Qian, Kodo Ito and Toshio Nakagawa
This paper considers and discusses analytically the optimal preventive maintenance (PM) policies of aged plants such as fossil‐fired power plants.
Abstract
Purpose
This paper considers and discusses analytically the optimal preventive maintenance (PM) policies of aged plants such as fossil‐fired power plants.
Design/methodology/approach
Shocks are assumed to occur at a nonhomogeneous Poisson process and the total damage due to each shock is additive. The system undergoes the PM at a certain time or the total damage exceeds a managerial level. The expected cost rate until PM is derived and optimal policies which minimizes it are discussed.
Findings
There exists a unique optimal time (T*) or managerial level (k*) which minimizes the expected cost rate. But there does not exist a positive pair (T*, k*), simultaneously.
Research limitations/implications
The damage occurrence distribution is assumed to be nonhomogeneous Poisson one.
Practical implications
Useful methods to consider the optimal PM policies for power plant engineers.
Originality/value
This paper contributes to users of aged power plants economically and practically.
Details
Keywords
Toshio Nakagawa and Qian Cun Hua
This paper considers a series‐parallel and a parallel‐series system, and investigates theoretically their fundamental characteristics: it is shown that the reliability of the…
Abstract
This paper considers a series‐parallel and a parallel‐series system, and investigates theoretically their fundamental characteristics: it is shown that the reliability of the series‐parallel system with the same number of series and parallel tends to one as its number goes to infinity. It would be of great interest that the golden ratio plays a role in analyzing two systems. Further, an optimal number of units for a series‐parallel system with complexity is derived.