Optimizing the maximum filling level of perfect storage in system with imperfect production unit

• Published in: Reliability engineering & system safety Vol. 225; p. 108629
• Main Authors: Levitin, Gregory, Xing, Liudong, Dai, Yuanshun
• Format: Journal Article
• Language: English
• Published: Barking Elsevier Ltd 01.09.2022; Elsevier BV
• Subjects: Algorithms; Interaction parameters; Mission success; Numerical analysis; Optimization; Preventive maintenance; Probabilistic models; Random repair time; Reliability engineering; Statistical analysis; Storage filling; Random repair time; Mission success; Storage filling
• ISSN: 0951-8320; 1879-0836
• DOI: 10.1016/j.ress.2022.108629

•A failure prone production system with a perfect storage is considered.•The production system can undergo maintenance and repair during its mission.•The influence of maximum storage filling on the mission success is analyzed.•An algorithm for the mission success probability evaluation is suggested.•The optimal storage filling problem is formulated and solved. Reliability of production systems with storage has recently attracted lots of research attentions. While the existing works have assumed certain maximum capacity of the storage, no models are available to examine the effects of the storage's maximum filling level Cmax on the mission success probability (MSP). This paper contributes by modeling two-sided effects of Cmax on the MSP of an imperfect production system subject to repairs and preventive maintenance (PM) during the specified mission time. In particular, a larger value of Cmax enables the storage to supply the system demand during a longer time when the production system is under repair or PM (enhancing the MSP); on the other hand, it leads to longer operation periods and consequently more frequent failures of the production system (reducing the MSP). To balance these conflicting effects, we formulate and solve the optimal storage filling problem, which determines the optimal value of Cmax to maximize the MSP. The optimization solution encompasses a new probabilistic model-based numerical algorithm proposed for the MSP evaluation. A case study of a water pump system is performed to demonstrate the influences of several system parameters and their interactions on the MSP and optimized Cmax.