The impact of autocorrelation on queuing systems
Management Science
Autocorrelation of cycle times in semiconductor manufacturing systems
WSC '95 Proceedings of the 27th conference on Winter simulation
Merging and splitting autocorrelated arrival processes and impact on queueing performance
Performance Evaluation
Computers and Electronics in Agriculture
Hi-index | 0.00 |
Autocorrelation has been pointed out as one of the most challenging issues in manufacturing systems modeling. Numerical experimentation has shown that it may either enhance or harm performance. Furthermore, there is not yet a general agreement in what a realistic autocorrelation model is or whether it is actually relevant for practical applications. This paper provides a simulation analysis of the effects on performance caused by manufacturing process parameters following autoregressive (AR) processes. AR time series are employed for modeling variations in parameters that happen at a time scale different from the corresponding to process cycle execution. Three basic configurations are analyzed: serial line, assembly process and a disassembly process. A case study from the natural slate tiles industry is presented showing the differences obtained in simulation results between a model in which independent and identically distributed (i.i.d.) assumptions are adopted and one in which autocorrelation effects are considered.