Models for Disassembly Lot Sizing Problem with Decisions on Surplus Inventory

Abstract

We consider a single-product Disassembly Lot Sizing Problem with Disposal (DLSPD) which is a problem arising in the context of disassembly systems. This is the problem of determining the quantity and time of the returned products to be disassembled while satisfying the demand of their parts or components over a planning horizon. Disassembly operation generates several components simultaneously. And, the demands are independent and not balanced which can generate unnecessary surplus inventory during planning horizon. Aggregate formulation (AGG) can be used to model this problem by considering disposal decision. Linear-Programming (LP) relaxation of this model doesn’t give very good lower bound, especially for the large size instances. We aim to improve lower bound of the problem. Facility Location-based formulation (FAL) and additional constraints (Valid Inequalities (VIs)) for the LP relaxation of AGG model are proposed. Computational results on generated test instances show that LP relaxation of FAL and AGG with additional constraints can obtain very strong lower bound within a very short computational time which is useful for the varied DLSPD (multi-level, multi-product, …).