Multi-objective optimization model and algorithm for hybrid parallel machine scheduling problem
Ya-ping FU,Min HUANG,Hong-feng WANG,Xing-wei WANG
DOI: https://doi.org/10.7641/CTA.2014.31233
2014-01-01
Abstract:A special hybrid machine scheduling problem is investigated since the combination of production processes can yield a special layout with the coexistence of serial and parallel production.In order to address this problem effectively,a mixed integer programming multiobjective model,which aims to minimize both the total flow time and the total number of tardy jobs,is built and an improved nondominated sorting genetic algorithm II(NSGA-II),where a heuristic method is proposed to generate and improve the quality and diversity of initial population,and a local search strategy is embedded to improve the quality and distribution of the achieved nondominated solutions,is designed in this paper.Compared with the classical multiobjective optimization algorithms,the experimental results on a number of stochastic examples with different scale show that the proposed model and algorithm have advantage in the convergence,spread and extreme value point,and can address effectively the investigated hybrid parallel machine scheduling problem.