Performance Prediction of a Series-parallel and Multi-Product Production Line with Unreliable Machines, Finite Buffers and Nonconforming Products

摘要

It is well known that performance prediction is extremely important to improve the flexibility of a series-parallel and multi-product production line with unreliable machines, finite buffers, and nonconforming products, but how to model and predict its performance is facing great challenges. A novel approximation iteration method based on a discrete state Markov chain and queueing model is developed. The equivalent machines processing rates in isolate state are determined by applying variability transition parameters via Markov chain firstly; Then the throughput of whole production line is determined according to the last workstations effective processing rate in steady-state, which is approximately predicted via queueing modelling; Thirdly, the unknown parameter throughput X((k))is computed by developing an approximate iterative algorithm procedure, then a modified queueing model is used to compute the performance. Finally, to assess the effectiveness of the proposed method, extensive numerical experimental results from the predictive approximation are compared to simulation models, which proves it accurate and believable.