Optimisation of Assembly Line Balancing Type-E with Resource Constraints Using NSGA-II

Abstract

Assembly line balancing of Type-E problem (ALB-E) is an attempt to assign the tasks to the various workstations along the line so that the precedence relations are satisfied and some performance measures are optimised. A majority of the recent studies in ALB-E assume that any assembly task can be assigned to any workstation. This assumption lead to higher usage of resource required in assembly line. This research studies assembly line balancing of Type-E problem with resource constraint (ALBE-RC) for a single-model. In this work, three objective functions are considered, i.e. minimise number of workstation, cycle time and number of resources. In this paper, an Elitist Non-dominated Sorting Genetic Algorithm (NSGA-II) has been proposed to optimise the problem. Six benchmark problems have been used to test the optimisation algorithm and the results are compared to multi-objective genetic algorithm (MOGA) and hybrid genetic algorithm (HGA). From the computational test, it was found NSGA-II has the ability to explore search space, has better accuracy of solution and also has a uniformly spaced solution. In future, a research to improve the solution accuracy is proposed to enhance the performance of the algorithm.

Optimisation of Simple Assembly Line Balancing Problem Type E: A Systematic Literature Review

“…This assumption has led to an increased use of assembly line resources. Jusop and Ab Rashid [47] consider ALB-E with resource constraints (ALBE-RC), such as the machine, tool, and worker required to complete an assembly process for a given product. In this work, three objective functions are investigated: the number of workstations, the cycle time, and the amount of resources.…”

Assembly line balance research methods, literature and development review

“…With the deepening of research on assembly line balancing problems, the complexity of the balancing problems has also continued to expand, and scholars have also begun to seek more intelligent algorithms to obtain satisfactory solutions in a short time. The application of different artificial intelligence algorithms on different balancing problems is the current research focus of scholars, because this type of algorithm has a good balance effect and can also be applied to large-scale problems, such as GA (Chen et al, 2001; Jusop and Rashid, 2016; Li et al, 2020; Liu et al, 2014, 2017; Mura and Dini, 2016; Pi et al, 2005; Simaria and Vilarinho, 2004; Tanhaie et al, 2020; Wang, 2006; Zacharia and Nearchou, 2012, 2013; Zhang et al, 2006), ACO (Mcmullen and Tarasewich, 2003), PSO (Liu et al, 2019; Xu et al, 2016; Zhang et al, 2016), SA (Mcmullen and Frazier, 1998; Suresh and Sahu, 1994; Vilarinho and Simaria, 2002), TS (Song and Han, 2002b), etc. Simaria and Vilarinho (2004) proposed an iterative GA for the mixed-model ALBP-II with parallel workstations and zoning constraints.…”

A linear physical programming model for assembly line balancing problem

“…The second group includes studies using metaheuristic approaches. Genetic algorithm (Gurevsky et al, 2012;Alavidoost et al, 2015;Mura and Dini, 2016;Jusop and Rashid, 2016;Zhang et al, 2020), artificial bee colony algorithm (Tang et al, 2016;Zhao et al, 2016;Zhang et al, 2018), ant colony optimization (Zheng et al, 2012;Samouei and Dezfoulian, 2017;Huo et al, 2018;Huang et al, 2020), discrete cuckoo search , hybrid genetic algorithm (Lin et al, 2009), simulated annealing (Dong et al, 2018;, memetic algorithm (Pereira et al, 2018), fish school search algorithm (De Albuquerque et al, 2016); migrating birds optimization metaheuristic (Janardhanan et al 2019), multi-objective evolution strategies (Yoosefelahi et al, 2012;Zacharia & Nearchou, 2016).…”