Modelling and optimization of distributed heterogeneous hybrid flow shop lot-streaming scheduling problem

Abstract

More enterprises are facing rapid and changing market demand and small but frequent orders, and they extend the production capability of manufacturing systems to respond to these changes. This paper studies a distributed heterogeneous hybrid flow shop lot-streaming scheduling problem (DHHFLSP) with the minimization of makespan. In the DHHFLSP, the processing capacity of each factory is different and each job can be split into several sub-lots. These sub-lots are assigned to several non-identical factories. The mixed-integer linear programming model (MILP) of DHHFLSP is established. To solve the DHHFLSP, eighteen constructive heuristics and an iterated local search algorithm (ILS) are designed. In constructive heuristics, the jobs are sorted according to several time-based heuristic rules or they are divided into several groups according to bottleneck stages, and then these jobs are assigned into factories through two NEH-based job assignment rules. In the ILS, the NEH-based heuristic (NEHafter) plus Largest medium rule is used to generate an initial solution. Two greedy insertion operators with or without critical factories are adopted to generate the perturbation solutions. Four greedy local search operators are designed to make a deep search. The influence of the parameters and main components are investigated by a comprehensive analysis. The comparisons with several related algorithms on extensive testing instances demonstrate the effectiveness and efficiency of the ILS algorithm.