Use this URL to cite or link to this record in EThOS:
Title: Integrated multi-objective optimisation of assembly sequence planning and assembly line balancing using particle swarm optimisation
Author: Ab Rashid, Mohd Fadzil Faisae
Awarding Body: Cranfield University
Current Institution: Cranfield University
Date of Award: 2013
Availability of Full Text:
Access from EThOS:
Access from Institution:
In assembly optimisation, Assembly Sequence Planning (ASP) and Assembly Line Balancing (ALB) optimisations currently performed in serial, present an opportunity for integration, allowing benefits such as larger search space leading to better solution quality, reduced error rate in planning and fast time-to-market for a product. The literature survey highlights the research gaps, where the existing integrated ASP and ALB optimisation is limited to a Genetic Algorithm (GA) based approach, while Particle Swarm Optimisation (PSO) demonstrates better performance in individual ASP and ALB optimisation compared to GA. In addition, the existing works are limited to simple assembly line problems which run a homogeneous model on an assembly line. The aim of this research is to establish a methodology and algorithm for integrating ASP and ALB optimisation using Particle Swarm Optimisation. This research extends the problem type to integrated mixed-model ASP and ALB in order to generalise the problem. This research proposes Multi-Objective Discrete Particle Swarm Optimisation (MODPSO), to optimise integrated ASP and ALB. The MODPSO uses the Pareto-based approach to deal with the multi-objective problem and adopts a discrete procedure instead of standard mathematical operators to update its position and velocity. The MODPSO algorithm is tested with a wide range of problem difficulties for integrated single-model and mixed-model ASP and ALB problems. In order to supply sufficient test problems that cover a range of problem difficulties, a tuneable test problem generator is developed. Statistical tests on the algorithms’ performance indicates that the proposed MODPSO algorithm presents significant improvement in terms of larger non-dominated solution numbers in Pareto optimal, compared to comparable algorithms including GA based algorithms in both single-model and mixed-model ASP and ALB problems. The performance of the MODPSO algorithm is finally validated using artificial problems from the literature and real-world problems from assembly products.
Supervisor: Tiwari, Ashutosh Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available