New models and efficient methods for single-product disassembly lot-sizing problem with surplus inventory decisions
Abstract
This paper addresses the problem of disassembly lot-sizing for the single-product type. Due to some specific characteristics of disassembly systems, surplus inventory can be generated while satisfying the demand for the components. Disposal decisions are considered here to avoid inventory accumulations throughout the planning horizon. Three new mixed-integer programming (MIP) formulations are proposed to model the problem. The formulations differ from each other concerning the quality of the lower bound provided by their linear relaxation, which is an important issue in MIP resolution methods. Two efficient heuristics are also investigated for real-case applications when MIP algorithms are not relevant. The three formulations and the performance of the heuristics are compared based on new randomly generated instances for disassembly lot-sizing problems. As a managerial insight, the disposal decisions in disassembly lot-sizing models are relevant to save inventory costs.