Two-sided based Material Allocation for Intermediate Inventory of Process Industry

Yiran Li,Ying Meng,Yuanqing Xia
DOI: https://doi.org/10.23919/CCC50068.2020.9189658
2020-01-01
Abstract:Two-sided matching is an effective decision methodology to solve resource allocation problem, aiming at satisfactions of two disjoint matching sides. Intermediate warehouse is designed to store semifinished products, which could ensure process industry logistics works continuously and flexibly. Allocating the surplus inventory materials to orders as a preprocess of production planning is a common method for made-to-order process manufacture. It relieves inventory pressure as well as improves material utilization. In this paper, a material-order two-sided matching model with consideration of both sides' preferences is formulated. Assessing both sides' satisfaction by complete and dynamic preference, the objective is to maximize the global matching preference and to get stable outcome. To solve the NP-hard problem efficiently, a tailored hybird algoritlun PDADE is proposed. Firstly, a preference deferred acceptance based heuristic is designed to obtain high quality initial solution. Then, a series of improved differential evolution strategies considering real number encoding of sparse two -dimension matrix is designed to achieve near-optimal solutions in short time for practical large scale instances. According to experiments of actual production data instances, the proposed algorithm outperforms the commercial solver and standard DE both on run time and quality of solutions.
What problem does this paper attempt to address?