Publication: Tamir Süresi Faz Tipi Dağılıma Uyan, Özdeş Olmayan İki Üniteli Bir Sistemin Kullanılabilirliğinin Analizi ve Simülasyonu
Abstract
Bu çalışmada, tamir süreleri Erlang, Hiper-Üstel ve Coxian dağılımına, çalışma süreleri de üstel dağılıma uyan özdeş olmayan iki üniteli modeller analiz edildi. Bu modellerde her ünite için üç durum vardır; çalışma, sağlam bekleme ve tamirdir. Modellerde, bir ünite çalışıyor ve diğeri sağlam olarak beklemektedir. Çalışan ünite arızalandığında diğeri hemen yerine geçer ve çalışmaya başlar. Arızalanan ünite ise tamir tesisinin ilk aşamasında tamir edilir. İlk aşamada tamir tamamlanmadıysa sonraki aşamada tamir tamamlanır. Eğer bir ünite tamir aşamasındayken diğer ünite de arızalanırsa bu durumda sistem çöker. Bahsedilen modeller Markov süreci ve Kolmogorov diferansiyel denklemleri yöntemleri ile oluşturuldu. Daha sonra limit dağılımları ile durum olasılıkları ve modellerin kullanılabilirliği elde edildi. Bu bağlamda, modellerin R programı yazılarak simülasyonu yapıldı. Sonuç olarak, nümerik değerler verilerek tam sonuçlar ile simülasyon değerlerinin birbirine yakın olduğu görüldü.
In this study, two non-identical unit cold standby systems with Erlang, Hyper-Exponential and Coxian distribution repair time and exponential distribution failure time were analyzed. These models have three state for each unit: working, cold standby and repair. In these models, one unit is working and the other is in standby without fail. When the working unit fails, the other immediately replaces and starts working. The failed unit is repaired at the first stage of the repair facility. If the repair is not completed in the first stage, the repair is completed in the next stage. If one unit is under repair and the other unit fails, then the system crashes. The mentioned models were created using the Markov process and Kolmogorov differential equations methods. Then, the limit distributions and state probabilities and the availability of the models were obtained. In this context, the models were simulated by writing the R program. As a result, numerical values were given and the simulation values were compared with the exact results.
In this study, two non-identical unit cold standby systems with Erlang, Hyper-Exponential and Coxian distribution repair time and exponential distribution failure time were analyzed. These models have three state for each unit: working, cold standby and repair. In these models, one unit is working and the other is in standby without fail. When the working unit fails, the other immediately replaces and starts working. The failed unit is repaired at the first stage of the repair facility. If the repair is not completed in the first stage, the repair is completed in the next stage. If one unit is under repair and the other unit fails, then the system crashes. The mentioned models were created using the Markov process and Kolmogorov differential equations methods. Then, the limit distributions and state probabilities and the availability of the models were obtained. In this context, the models were simulated by writing the R program. As a result, numerical values were given and the simulation values were compared with the exact results.
Description
Citation
WoS Q
Scopus Q
Source
Volume
Issue
Start Page
End Page
77
