This paper introduces an approach to enhance production scheduling systems through the implementation of an artificial hormone algorithm. It focuses on lots that must complete several processing stages across various machines, with each lot’s specific sequence of stages defined by a pre-determined recipe. The system’s flexibility is limited to selecting machines for the upcoming stage and altering the processing sequence of lots waiting at a machine. Optimizing this type of job-shop scheduling is recognized as an NP-hard challenge. The proposed method utilizes artificial hormone mechanisms to signal the priority of specific lots and the demand for new lots at certain machine types, thereby leveraging local information for optimization purposes. The findings suggest that employing the artificial hormone system leads to approximately a 5% performance enhancement compared to traditional First Come-First Serve strategies.
Wilfried Elmenreich, Alexander Schnabl and Melanie Schranz
Institute of Networked and Embedded Systems, Alpen-Adria-Universität Klagenfurt, Klagenfurt, Austria
Lakeside Labs, Klagenfurt Austria
