A comparative study of multi-objective algorithms for the assembly line balancing and equipment selection problem under consideration of product design alternatives

A realistic and accurate product cost estimation is of high importance during the design phases of products and assembly lines. This paper presents a methodology that aims at supporting decision makers during the design phases of assembly lines by taking into consideration product designs, processes and resources alternatives. First, we introduce a new variant of the Assembly Line Balancing and Equipment Selection Problem, in which Product Design Alternatives are considered. Since the ability to estimate product costs provides grounds for making better decisions, a new detailed cost model whose aim is to translate the complex and interrelated consequences of product design and manufacturing technologies and processes choice into one single cost metric is proposed. In order to solve the problem under study, 34 Multi-Objective Algorithms were developed. The list of developed algorithms includes variants of Evolutionary Algorithms, Ant Colony Optimisation, Artificial Bee Colony, Cuckoo Search Optimisation, Flower Pollination Algorithm, Bat Algorithm and Particle Swarm Optimisation. The performances of all these algorithms are compared based on fifty well-known problem instances in accordance with four multi-objective quality indicators. Finally, the algorithms are ranked using a nonparametric statistical test