Abstract: An improved discrete krill herd (IDKH) algorithm is proposed for the complex product assembly scheduling problem. The objective is to minimize the makespan. An assembly scheduling model is established by analyzing the characteristics of a complex product process flow. The transformation between scheduling solution and population individual is realized by using permutation-based coding and heuristic-based decoding methods. A local search and a restart operation procedure are presented to improve the exploitation and global exploration ability of basic krill herd (KH). The parameters of the proposed IDKH are calibrated by using a design of experimental approach. And a comparative evaluation is conducted with the well-known algorithms. The results show that the proposed IDKH has advantage over genetic algorithm, estimation of distribution algorithm, gravitational search algorithm, and basic KH in terms of quality and stability. Key

Key words: krillherdalgorithm, assemblyscheduling, complexproduct, hybridflow-shopscheduling