Batch sizing with random yield and imperfect inspection and process compressibility

This paper presents a new model for batch sizing in an imperfect production system that utilizes from an inspection subsystem for screening defective items from non-defectives. The needed setup time for starting the rework process is considered as a stochastic parameter. Regular production process and rework process could be done with different production rate and different stochastic percent of defectives. In this way, batch sizing is accompanied with random yield. Also production and / or rework processes could be performed in a time interval for each item by paying its corresponding cost. On the other hand, inspection is involved with some errors that result in returned items from the customers back to the producer. The proposed model considers many realistic conditions to obtain an optimal production plan of some decision variables like batch size, backlog, regular production rate, and rework rate simultaneously. Finally the proposed model is solved by the PSO algorithm and computational experiences are reported.