Using constraint-satisfaction to optimise disassembly sequences generated from AND/OR information

Abstract

This paper applies a constraint-based method, shown to successfully find the optimal order of component removal using precedence information, to finding the optimal sequence via AND/OR data. Development of disassembly theory is vital for practically all product end-of-life strategies. In particular, Disassembly Process Plans (DPPs) allow the careful selection and removal of components for recycling and reuse. While AND/OR graphs explicitly embody all of the feasible DPPs of an assembly, generating the best plan is non-trivial due to the rapid increase in the number of solution paths. In contrast, defining the physical restrictions between the components using precedence relationships is implicit and hence compact. By converting AND/OR representations into precedence ones, not only is the size of the problem reduced, modified techniques developed previously are applicable. Dependent upon the structure of the subassemblies, two different approaches are used to achieve this: the introduction of dummy components, and by means of the set of operations upon the subassemblies. Thus an all-encompassing methodology is created that can deal with any DPP representation such that the optimal sequences may be found.

Keywords:

• constraint programming
• optimisation
• disassembly
• AND/OR graphs
• precedence relations

Acknowledgements

This study was supported, in part, by a Grant-in-Aid for Forming Strategic Research Infrastructure from the Ministry of Education, Culture, Sport, Science, and Technology, Japan (MEXT), 2008–2012 (S0801058).