Non-stationary models of manufacturing systems: relevance and analysis

Abstract

Performance evaluation studies in manufacturing systems have traditionally considered models in which the arrival process and service process are time independent. Real-world manufacturing systems however, are subjected to highly complex and usually time-dependent input workloads. This motivates the study of performance models of manufacturing systems under nonstationary conditions. In this paper, we present several situations in manufacturing systems where nonstationary models are relevant. For studying such models, transient analysis is more appropriate than steady-state analysis. We explore various techniques for analyzing such models, including numerical and simulation techniques, and present two illustrative examples.