Back to results list
Show full item record
Please use this identifier to cite or link to this item:
|Title:||Integrated planning of berth allocation and quay crane assignment problems||Authors:||Ma, Hoi Lam||Degree:||Ph.D.||Issue Date:||2015||Abstract:||Terminal operation efficiency is regarded as the most critical issue in sea-freight transportation network. Therefore, the ultimate objective of this research study is to improve the efficiency in terms of the total operating cost and total servicing time required, meanwhile the quay space utilization. Accordingly, three main research questions have been addressed, i) Defining customer importance and maximizing customer service level, ii) Optimizing quay crane assignment and utilization, and iii) Dealing with multi-continuous berth layout. The thesis starts with Berth Allocation Problem (BAP) because it directly influences the customer service level represented by vessel waiting time, handling time, and completion time. To survive in the rigorous competitive environment nowadays, terminals strive to retain their customers by providing good service quality especially to those important ones. However, in literature customer importance is usually by using defined either by i) customer relationships or ii) handling volume. Both approaches have some drawbacks. Therefore, this thesis proposed a new defining approach to consider both factors simultaneously. Accordingly, a new Genetic Algorithm for BAP (GA-BAP) is proposed. Experimental results demonstrated that the proposed approach can serve many customers with good relationship, meanwhile maintaining a high handling volume. To improve terminal operations efficiency, optimization of quay cranes assignment is one of the key issues. Quay Crane Assignment Problem (QCAP) and BAP are highly interrelated. Therefore, integrated planning of berth allocation and quay crane assignment has been studied. However, holistically solving this problem is very complicated. Thus, decomposition approach is proposed. Moreover, a new methodology named Two-level Genetic Algorithm (TLGA) is proposed. Furthermore, for better Quay Crane (QC) utilization, variable-in-time quay crane assignment is further studied. In literature, time dimension is usually in hourly based. However, in transshipment terminal, vessel staying time is usually short (usually a few hours) comparing to traditional gateway terminals. Thus, hourly based approach may cause QC idling significantly. Therefore, a 15-minute based approach is proposed as suggested by industrial practitioners. A novel Three-level Genetic Algorithm (3LGA) with QC shifting heuristics is proposed to fulfill the research gap. The results show that the proposed 15-minute approach reduces QC idling remarkably. Meanwhile, the proposed 3LGA far outperforms the traditional GA in this problem. Lastly, a novel multi-continuous berth layout is studied driven by the practical needs in real wharf layout. In literature, many researchers usually applied discrete or hybrid berth layout modeling approach. However, this induces low space utilization. Therefore, some researchers proposed continuous berth layout model. However, such modeling approach currently cannot be applied to berth layout with discontinuity. Accordingly, a novel Mixed Integer Programming approach is proposed. To bring further the model close to reality, yard storage assignment planning has also been considered. Ultimately, an integrated BAP with variable-in-time QCAP and yard storage assignment in the multi-continuous berth layout is studied in this thesis. A Guided Neighbourhood Search (GNS) is developed to improve the optimization performance. Experiment results demonstrated that the proposed model can remarkably reduce the total terminal operating cost and computational time significantly.||Subjects:||Marine terminals -- Management.
Harbors -- Management.
Hong Kong Polytechnic University -- Dissertations
|Pages:||xvii, 180 pages : illustrations|
|Appears in Collections:||Thesis|
View full-text via https://theses.lib.polyu.edu.hk/handle/200/8299
Citations as of May 22, 2022
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.