Deadlock-free scheduling of an automated manufacturing system using an enhanced colored time resource Petri-net model-based evolutionary endosymbiotic learning automata approach

Deadlock-free scheduling of parts is vital for increasing the utilization of an Automated Manufacturing System (AMS). An existing literature survey has identified the role of an effective modeling methodology for AMS in ensuring the appropriate scheduling of the parts on the available resources. In this paper, a new modeling methodology termed as Extended Color Time Net of Set of Simple Sequential Process with Resources (ECTS3PR) has been presented that efficiently handles dynamic behavior of the manufacturing system. The model is subsequently utilized to obtain a deadlock-free schedule with minimized makespan using a new Evolutionary Endosymbiotic Learning Automata (EELA) algorithm. The ECTS3PR model, which can easily handle various relations and structural interactions, proves to be very helpful in measuring and managing system performances. The novel algorithm EELA has the merits of both endosymbiotic systems and learning automata. The proposed algorithm performs better than various benchmark strategies available in the literature. Extensive experiments have been performed to examine the effectiveness of the proposed methodology, and the results obtained over different data sets of varying dimensions authenticate the performance claim. Superiority of the proposed approach has been validated by defining a new performance index termed as the ‘makespan index’ (MI), whereas the ANOVA analysis reveals the robustness of the algorithm.