Session Aims & Scope
A groundbreaking industrial revolution has been recently instigated in the traditional shipping sector by a new wave of information technologies encompassing the Internet of Things, artificial intelligence, cloud computing, and big data. Successive strategies have emerged within the shipping domain, aiming to steer manufacturing and shipping operations toward an intelligence-driven and service-oriented paradigm. In addressing this challenge, the deployment of Digital Twin （DT） technology within the ship and offshore facilities industry emerges as a potent strategy to enhance design efficacy, elevate digital operational controls, and foster the progression of ancillary industries. Across all lifecycle phases, DT technology amalgamates physical, informational, and cross-domain twin data, propelling a service framework capable of delivering interconnected intelligent services to all participants in the ship industry spectrum.
To share the latest advances in DT technologies for shipbuilding, the Digital Twin International Conference in 2024, is extending a Special Session for Digital Twins for Shipbuilding. The proposed Special Session particularly fits the following topics, but are not limited to:
- Digital Twin Ship Design and Optimization
- Digital Twin Ship Modeling and Simulation
- Digital Twin Ship Assembly and Optimization
- Digital Twin Shipbuilding Workshop Scheduling
- Digital Twin Shipbuilding Equipment Maintenance
Session Chair(s)
Chair
Honggen Zhou
Professor, Jiangsu University of Science and Technology (China)
hgzhou205@just.edu,cn
Co-Chair
Jinfeng Liu
Associate professor, Jiangsu University of Science and Technology (China)
liujinfeng@just.edu.cn
Co-Chair
Vsevolod V. Koryanov
Associate professor, Moscow State Bauman Technical University (Russia)
vkoryanov@bmstu.ru
Co-Chair
Xuwen Jing
Professor, Jiangsu University of Science and Technology (China)
Jingxuwen_just@163.com
Session Presentation
1.
Bo Zhou
Professor
Dalian University of Technology (China)
Title: Research on Digital Twin of Self elevating Platform
Abstract
On the basis of the initial equipment of the offshore self elevating platform, this study adds a series of sensor systems to monitor structural data and ensure the information transmission and reception functions of key monitoring equipment. Based on the completion of the monitoring system, develop a software system database and 3D display module, establish a comprehensive data acquisition and simulation calculation plan, build a digital 3D display platform for the platform structure, and realize the digital twin and intelligent application of the self elevating platform.
2.
Changhui Liu
Associate Professor
Tongji University (China)
Title: Optimal Design of Fixture Layout for Compliant Part with Application in Ship Hull Assembly Process
Abstract
In the ship assembly process, a large number of compliant parts are involved. The ratio of the part thickness to the length or the width is very small. Fixture design is a critical task in the ship assembly process due to its impact on the deformation and dimensional variation of the compliant parts. The current practice in the ship industry, the fixtures are uniformly distributed, which is non-optimal and large deformation will occur. This speech will talk about a series of studies of Prof. Liu for optimal design of fixture layout in the ship assembly process by integrating direct stiffness method and Metaheuristic algorithm, which significantly reduced the deformation of the compliant part in the ship assembly process.
3.
Shun Wang
Associate Professor
Dalian Maritime University (China)
Title: Forming Difficulty Evaluation of Line Heating for Complex Curved Ship Plate
Abstract
To solve the problem that the lack of reference for hull plate division leads to the difficulty and low efficiency of curved plate forming, the forming difficulty evaluation of complex curved hull plate is researched. Considering the characteristics in curved plate forming process, the weight coefficient of different affecting factors on the forming difficulty of curved plate is quantified. The forming difficulty of curved hull plate is calculated. By comparing and analyzing the forming evaluation results and forming time data of curved hull plate, the rationality of the evaluation method is verified. The research can provide reference for the reasonable division of hull plate.
4.
Cheng Luo
Associate Professor
Shanghai Jiao Tong University (China)
Title: Digital Twin Model-driven Adaptive Welding of Curved Ship Blocks
Abstract
Welding path planning is vital to improve the efficiency of the robot system in welding of curved ship blocks. Common welding path planning methods is mainly depending on discrete points fitting to determine the curved seam positions in physical welding conditions, leading in both decreases of efficiency and accuracy. In this report, a deep-learning empowered-digital twin system was proposed to plan and adjust welding paths by modeling the seams positions and features during welding process. The model successfully facilitated the adaptive control of weld positions for curved seams in real time, resulting in the improvement of welding quality.
5.
Vsevolod V. Koryanov
Associate professor
Moscow State Bauman Technical Universit (Russia)
Title: Digital Twins for Modeling Flight Environment Processes in Aerospace Spacecraft Design
Abstract
Currently, the aerospace industry is developing in almost all areas: these include near-Earth constellations of spacecraft, and the development of deep space exploration. Designing spacecraft that move in deep space is impossible without the use of digital twins. The process is complicated by the fact that it is often very difficult and expensive or impossible to create full-fledged experimental setups that simulate the process of spacecraft movement in space. The processes that occur when spacecraft move in space are sometimes very fleeting and important. In the direction of designing spacecraft, the use of digital twins is very relevant and necessary.
6.
Xiaolei XIANG
Associate professor
Curtin University WASM (Australia)
Title: Deep Reinforcement Learning for Optimizing fleet Management in Shipbuilding: A Comparative Study
Abstract
This research investigates the application of Deep Reinforcement learning (DRL) algorithms to optimize fleet management in shipbuilding operations. Three DRL models-modified Actor-Critic (AC), Prioritized Experience Replay (PER) Dueling Double Q-network (DDQN), and Proximal Policy Optimization (PPO)-are developed and accessed for their effectiveness in improving efficiency and productivity. The study compares their DRL approaches to traditional methodologies, demonstrating superior performance in enhancing operational efficiency. The findings provide valuable insights into the potential of DRL algorithms to drive intelligent fleet management and resource optimization in the shipbuilding industry.
7.
Lu Wang
Assistant Professor
City University of Hong Kong (China)
Title: Modeling the laser welding during ship building
Abstract
Metal melting and solidification is the basic process in both laser powder bed fusion and laser welding during ship building and repairing. However, the quality of the fabrication is difficult to be controlled though trail-and-error experiments due to the laser influencing factors. The high cost of experimental measurements also poses difficulties for the processing parameter optimization. To solve this tough situation, high-fidelity simulation is an effective approach to explore the mechanisms behind it and reduce the defects formation. In this report, a multi-physics thermal-fluid model, incorporating heat transfer, metal evaporation, Marangoni effect, etc., is brought up to simulate the pore formation. To optimize the process, an external magnetic field is applied to generate high intensity Lorentz force to modify the molten pool flow and adjust the grain morphology. These simulation results have been validated with advanced experiment measurements, which lays a solid foundation for further study to optimize the welding process.
8.
Wenjing Xing
Assistant Professor
Yamagata University (Japan)
Title: Deep Learning Analysis of Spray Particles Characteristics for Gas Turbine Engine
Abstract
This study aims to comprehensively investigate the atomization process and characterize the spray, encompassing both ligaments and droplets, through the application of deep learning techniques. The spray, injected from a twin-fluid atomizer into a crossflow, is analyzed using a hybrid approach involving CNN-based deep-learning and image processing to detect spray particles as regions of interest (ROI). Post-analysis of the obtained ROIs involves assessing displacement, sizes, and posture (orientation) distribution of spray particles. Additionally, a novel definition is proposed to determine the average velocity vectors of spray particles. The results reveal a noticeable change in the flying direction, transitioning from predominantly horizontal angles to nearly vertical angles, especially evident with an increasing atomizing air flow rate.