A Chemical Queueing Framework for Bulk Service Systems with Rework, Starting Failure and Server Vacations: An Adaptive Neuro-Fuzzy Inference System Approach

This paper describes a chemical queueing framework for modeling bulk service systems that include important operational aspects of starting failure, rework, inspection, repair, and a number of vacations. By applying queueing theory to chemical and pharmaceutical production situations, the model accounts for the real-time dynamics of batch processing systems compared to the deterministic approaches. Using the supplementary variable technique, steady-state analysis was used to identify stability criteria. Additionally, a cost model and a few performance metrics are developed. Numerical illustrations are provided to help visualize the impact of various characteristics of the system. In addition, the Adaptive Neuro-Fuzzy Inference System (ANFIS) is used for the validation and prediction of the system behavior.