• Proceedings of the Korean Operations and Management Science Society Conference
• /
• 2005.05a
• /
• pp.348-352
• /
• 2005

본 논문은 조선소에서 계획기간동안 각 플랜트의 생산일정에 따른 블록의 모든 수송요구량이 미리 알려져 있는 정적인(static) 수송환경을 고려한 트랜스포터 일정계획문제를 다룬다. 조선소 내에서 블록들의 수송은 몇 가지 특성을 가진다. 500톤을 초과하는 거대 블록들은 2대 이상이 결합된 새로운 형태의 트랜스포터를 이용하여 운반된다. 운반이 요구되는 블록들 중 일부는 계획기간동안 규정된 출발요구시각에 pick-up되기를 요구하는 반면 나머지 블록들은 계획기간동안 규정된 도착요구시각 전에 delivery되기를 요구한다. 이러한 트랜스포터의 결합 및 시간 제약은 문제를 더욱 복잡하게 하는 요인이 된다. 본 논문에서는 계획기간 내에 모든 블록운반 요구사항들(출발지와 도착지, pick-up과 delivery 시각, 톤수, 결선 등)을 만족시키는 트랜스포터 일정계획문제를 다룬다. 트랜스포터 일정계획문제에서 중요한 주제는 총 로지스틱스 시간 (공차운행시간, delay 시간 및 tardy 시간)을 최소화하기 위해 계획기간 내에 최소의 트랜스포터 운영대수로 각 트랜스포터에 어떤 블록을 할당하고 어떤 순서로 운반할 것인 지를 결정하는 것이다. 이에 본 논문에서는 주어진 트랜스포터 일정계획문제를 해결하기 위해, 1단계에서 최소비용 네트워크 흐름문제를 근간으로 트랜스포터의 최소사용대수를 구하고 이를 기초로 2단계에서 각 트랜스포터에의 블록 할당과 운반순서를 결정하는 2단계 휴리스틱 알고리즘을 제안한다.

• Journal of Navigation and Port Research
• /
• v.38 no.3
• /
• pp.299-305
• /
• 2014

In a ship-building, a transporter is the moving equipment to transport ship blocks from a workshop to another in a shipyard. The efficient scheduling of transporters has an important role for the operation of a building ship to be completed on schedule. There are the previous studies on the transporter scheduling for moving blocks in a shipyard. These studies have no consideration for the transporter combination to increase the productivity of moving blocks. This paper presents an efficient transporter scheduling model considering transporter combination explicitly. The objective of this model is to minimize the operational cost and maintain the workload balance among transporters. we also present three heuristics algorithms based on tabu search for finding the solution of the model. The efficiency of the proposed heuristics are verified with several computational tests.

• Journal of the Korea Society of Computer and Information
• /
• v.28 no.12
• /
• pp.49-56
• /
• 2023

In this study, we propose a genetic algorithm (GA) to optimize the allocation and operation order of transporters. The solution in the GA is represented by a set of lists each of which the operation order of the corresponding transporter. In addition, it was implemented in the form of a hybrid genetic algorithm combining effective local search operations for performance improvement. The local search reduces the number of operating transporters by moving blocks from a transporter with a low workload into that with a high workload. To evaluate the effectiveness of the proposed algorithm, it was compared with Multi-Start and a pure genetic algorithm through a simulation environment similar in scale to an actual shipyard. For the largest problem, compared to them, the number of transporters was reduced by 40% and 34%, and the total task time was reduced by 27% and 17%, respectively.

• Journal of the Society of Naval Architects of Korea
• /
• v.54 no.3
• /
• pp.187-195
• /
• 2017

The operation planning of transports used to move blocks is the one of key factors. Furthermore, reducing the running time through the effective plan contributes to pulling forward the whole logistic process of the shipyard and substantially saving the fuel consumption of itself as well. The past researches of the transporter focused on finding only the shortest distances, so called, Manhattan distance. However, these searching approaches cannot help having the significant difference in the real operational time and distance with the minimum cost approach which considers the speed retardation for turns or safety at the intersection. This study suggests the noble transporter's operational model which could take account of the consuming operational time around the crossroads on the shipyard. Concretely, the proposed method guarantees the minimization of transporters' turns and passage number which are huge burdensome to the operation time and the whole planning of transports with the given period. Resultantly, this paper is willing to explain the appropriateness of our approach, compared with the previous ones.

• Korean Journal of Computational Design and Engineering
• /
• v.21 no.2
• /
• pp.177-185
• /
• 2016

In this paper, an optimal planning system for operating transporters in shipyard is developed. The system is designed to utilize the geometries of shipyard, and manage the data of blocks and transporters directly. There are four major menus such as shipyard map management based on GIS, block transportation request, transporter management, and optimal transportation planning in the system. The geometries and properties of the shops, roads, and addresses are manipulated in the shipyard map management menu. The block transportation requests and the properties of transporters are managed in the block transportation request and transporter management menus, respectively. The optimum transportation is planned automatically for minimizing the unload times of the transporters, and the optimum transportation plans are confirmed and printed to the transporter drivers. The effectiveness of the system was verified through the application to a large-sized shipyard.

• Journal of the Society of Naval Architects of Korea
• /
• v.54 no.5
• /
• pp.421-429
• /
• 2017

It is very important to manage the position of the blocks in the shipyard where the work is completed, or the blocks need to be moved for the next process operation. The moving distance of the block increases according to the position of the block stockyard. As the travel distance increases, the number of trips and travel distance of the transporter increases, which causes a great deal of operation cost. Currently, the selection of the block position in the shipyard is based on the know-how of picking up a transporter worker by the production schedule of the block, and the location where the block is to be placed is determined according to the situation in the stockyard. The know-how to select the position of the block is the result of optimizing the position of the block in the shipyard for a long time. In this study, we used the accumulated data as a result of the operation of the yard in the shipyard and tried to select the location of blocks by learning it. Decision tree learning algorithm was used for learning, and a prototype was developed using it. Finally, we prove the possibility of selecting a block stockyard through this algorithm.