A multi-objective evolutionary algorithm for the hybrid flowshop rescheduling problem with lot streaming

With random disturbance events, the actual workshop environment is often dynamic, which makes the original schedule worse or even unfeasible. In this paper, considering the random disturbance of machine breakdown, a multi-objective hybrid flowshop rescheduling problem with lot streaming (MOHFRP_LS) is described. Firstly, a multi-objective optimization model is established with the total flow time and total starting time deviations of sublots. Secondly, based on the the technique for order preference by similarity to an ideal solution (TOPSIS) is introduced to evaluate the solution fitness, a multi-objective migrating birds optimization algorithm (MMBO) is proposed. In the algorithm, a dynamic decoding strategy by considering machine breakdown is proposed, and an improved mechanism is proposed to further improve the solution quality. Population initialization based on Glover operation is carried out to utilize the information contained in the original schedule. In the competition mechanism, population rearrangement technology and fast non-dominant sequencing technology are introduced to adjust the V-shaped population structure, and population reproduction strategy was used to carry out secondary updating to improve the population quality. In the scout mechanism, a local search approach is proposed based on Pareto property to improve the quality of alternative solutions. Finally, the effectiveness of MMBO is verified by experimental comparison and analysis.