Reliability and availability analyses of multi-failure renewal system under waiting
Loading...
Date
2020-09
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
G.B. Pant University of Agriculture and Technology, Pantnagar - 263145 (Uttarakhand)
Abstract
Due to the complexity of the real life systems it becomes mandatory to evaluate their reliability and availability. In most emerging repairing complex models, failures are not self announcing, they are found to be hidden/unrevealed. Keeping this fact in the view, the focus of the present research is on the development of reliability and availability models with the help of derived preposition, Markov process, Gumbel-Hougaard copula and ANN. In this study, six models are introduced. Model 1 deals with the methodical prepositions on availability and maintenance cost of competing risk system under periodic inspection and corrective repair,
which follows (M/M/1) : (/ FCFS) queue model. Analytical results on these prepositions such as point availability, long-run availability and maintenance cost along with sensitivity with respect to different inspection periods have been examined. Model 2 discusses some prepositions on availabilities of multi failure modes system under three factors: age-based inspection policy, maintenance strategies including preventive and corrective repair and (M/Mc /1) : (/ FCFS) , where Mc represents Gumbel-Hougaard copula which couples two different distributions: general and exponential distribution. Model 3 focuses on the prepositions of system availability under semi vacation policy named as working vacations and (M/M/1) : (/ FCFS) queue
model. This model studies periodic examinations and perfect corrective maintenance modeling of a multi failure modes system following (M/M/1):(/FCFS) queue model and the working vacations during repair. As the jth (j=1,2,…, β) failure occurs with failure rate λj, respective corrective repair with repair rate μj having random time Wj, (j=1,2,…, β) is carried out. Model 4 studies age-based periodic examination and perfect maintenance modeling (including PR and CR) of a multi failure modes system following (M/MGH /1) : (/ FCFS) queue model in repair. As the jth (j=1,2,…, β) failure occurs with failure rate λj, respective corrective repair with repair rate μj follows a new queuing model defined as (M/MGH /1) : (/ FCFS) , where Mc represents Gumbel-Hougaard copula which couples two different distributions: general and exponential distribution. Furthermore, the generalization of the classical vacation queue is introduced during repairing which takes random time Wj, (j=1,2,…, β). Some methodical prepositions are derived on point availability and long-run availability for the considered complex system. Model 5 investigates the reliability assessment and reliability improvement of cloud computing system using stochastic process, supplementary variable technique and Artificial Neural Network. In this model, the proposed system failure occurs due to different parameters and repair action takes place with two different distributions: general and exponential which are coupled by Gumbel-Hougaard copula. This study is estimated by embedding Markov process using Laplace transforms to estimate the system state probabilities, reliability, sensitivity and mean time to failure. Model 6 discussed in two sections. Section [1] examines Model 6A which studies many reliability characteristics such as transition state probabilities, reliability, MTTF and sensitivity of power transformer with dissimilar parameters as insulating oil and insulation paper using the Laplace transforms supplementary variable technique (SVT) and Gumbel-
Hougaard copula. Reliability measures of the transformer consisting of both the parameters: insulating oil and paper following two kinds of repair combining Gumbel-Hougaard family of copula have been determined. Section [2] corresponding to Model 6B deals with the analysis of reliability of transformer with the help of stochastic process via SVT and Neural Network Approach. The failure of the mentioned system occurs due to parameters of insulating paper and oil. The rates of repair for both insulation oil and insulation paper follow general distribution and insulation oil repair rates follow exponentially distribution. Also, whenever two types of repair are feasible, it has been coupled through Gumbel-Hougaard copula. Various sorts of failures and repair rates are treated as weights while investigating with the help of neural network approach which are followed by exponential distribution. System state probabilities, reliability, up and down states probabilities have been assessed for the model by using Markov process using Laplace transforms. To demonstrate the study with illustrations, numerical examples are also included in each proposed model.