Unified Colored Timed Petri net models for the adaptive control of disassembly systems

A method is presented for the robust control of a disassembly system that undergoes process plan variations due to heterogeneity of demands and product conditions. Two colored timed Petri net (CTPN) models is proposed and extended for the hierarchical modeling of both the system architecture and disassembly plans corresponding to the products to be processed. To address the high levels of uncertainty inherent in the operation of the disassembly system, a hybrid Bayesian network (HBN) is adopted to intercommunicate with the two CTPN models, allowing the system to exploit any newly available information for future decision-makings. An algorithm is then introduced to maximize the expected returns from the disassembly process while efficiently utilizing the system resource to fulfill order requirement.