Search

In multi-manned assembly lines, more than one operator is allowed on the same workstation. Limited research has been carried out on sequence-dependent set-up times. Therefore, this article optimizes the multi-manned assembly line balancing problem with sequence-dependent set-up times using an enhanced migrating birds optimization (EMBO) algorithm. Sequence-dependent set-up times are embedded in two ways: forward and backward set-ups. EMBO designs a mixed neighbourhood search and acceptance criteria to help it escape from local optima, and a diversification mechanism to enhance the global exploration ability. Three types of experiment and two scenarios, with low and high set-up times, are conducted to test the performance of the mixed-integer linear programming (MILP) model and EMBO. The first experiment compares three MILP models (no, low and high set-up times) to illustrate the significance of set-up times. The final two experiments demonstrate the effectiveness of EMBO by comparison with the MILP model and six state-of-the-art methods. [ABSTRACT FROM AUTHOR]

Steelmaking and the continuous-casting (SCC) scheduling problem is a realistic hybrid flow shop scheduling problem with continuous-casting production at the last stage. This study considers the SCC scheduling problem with diverse products, which is a vital and difficult problem in steel plants. To tackle this problem, this study first presents the mixed-integer linear programming (MILP) model to minimize the objective of makespan. Then, an improved migrating birds optimization algorithm (IMBO) is proposed to tackle this considered NP-hard problem. In the proposed IMBO, several improvements are employed to achieve the proper balance between exploration and exploitation. Specifically, a two-level decoding procedure is designed to achieve feasible solutions; the simulated annealing-based acceptance criterion is employed to ensure the diversity of the population and help the algorithm to escape from being trapped in local optima; a competitive mechanism is developed to emphasize exploitation capacity by searching around the most promising solution space. The computational experiments demonstrate that the proposed IMBO obtains competing performance and it outperforms seven other implemented algorithms in the comparative study. [ABSTRACT FROM AUTHOR]