Modelling and optimisation of assembly line balancing problem with resource constraint

• Main Author: Nur Hairunnisa, Kamarudin
• Format: Thesis
• Language: English
• Published: 2019
• Subjects: TJ Mechanical engineering and machinery
• Online Access: http://umpir.ump.edu.my/id/eprint/30318/1/Modelling%20and%20optimisation%20of%20assembly%20line%20balancing%20problem%20with%20resource%20constraint.wm.pdf; http://umpir.ump.edu.my/id/eprint/30318/
• Institution: Universiti Malaysia Pahang

Assembly Line Balancing (ALB) is about distributing the assembly tasks into workstations with the almost equal workload. Previous research mostly assumed that all workstations are having similar capabilities including the machines, tools and worker skills. Recently, researchers started to consider the resource constraints in ALB such as machine and worker. Optimisation of ALB with resource constraints gives a huge benefit to the industry such as increase line efficiency, optimise the resources utilisation and can reduce production cost. This research presents Assembly Line Balancing with resource constraints (ALB-RC) for a simple model with the objectives to minimise the workstation, machine and worker. For the optimisation purpose, this research introducesGenetic Algorithm (GA) with two new crossovers. The crossovers are developed using a ranking approach and known as rank-based crossover type I and type II (RBC-I and RBCII). The GA with new crossover is tested against popular combinatorial crossovers with a wide range of problem difficulties consisting of 17 benchmark problems. The performance of the proposed GA with new crossover in optimisation ALB-RC is finallyvalidated using an industrial case study. The computational experiment results indicated that the proposed GA with new crossovers are able to find the optimal solution for ALBRC better than popular combinatorial crossovers. Meanwhile, the results of industrial case study validated that the proposed ALB-RC model is capable to be used for the realindustrial problem. At the same time, the result indicated that the GA with rank -based crossover is capable to optimise real-life problem. As a comparison, the number of workstation, machine/tools and workers had reduced between 10 – 15% for the optimisedlayout using GA with RBC, compared with the original layout in the case study problem.