A novel approach for motion predictions of a semi-submersible platform with neural network
- PDF / 3,064,050 Bytes
- 13 Pages / 595.276 x 790.866 pts Page_size
- 16 Downloads / 253 Views
ORIGINAL ARTICLE
A novel approach for motion predictions of a semi‑submersible platform with neural network Yanfei Deng1,3 · Wei Feng3 · Shengwen Xu2 · Xiqia Chen3 · Bo Wang3 Received: 12 August 2019 / Accepted: 27 July 2020 © The Japan Society of Naval Architects and Ocean Engineers (JASNAOE) 2020
Abstract A neural-network-based prediction of motion responses of a semi-submersible is presented here. The fully connected neural networks and the long–short-term memory networks were employed to establish the neural networks for motion predictions. The effects of network architectures and time steps were investigated in depth. The predicted results were compared with the measurements and the predictions based on other conventional methods. The results demonstrate that the neuralnetwork-based approach could offer a fast and accurate prediction of heave, roll, and pitch responses of a semi-submersible. This provides a promising alternative for evaluations of hydrodynamic performances of new designs and monitoring of the dynamic behavior of in-service floating structures. Keywords Neural network · Motion prediction · Semi-submersible
1 Introduction Marine structures play an essential role in the exploitation and utilization of marine resources by providing adequate spaces for offshore operations and personnel residences, carrying various facilities, equipment, and resisting harsh environments. Among them, the semi-submersible platform is one of the widely used floating structures for its excellent hydrodynamic performances. The motion predictions of semi-submersibles are of great importance, which are directly related to the safety of structures and life, operation rates, etc. For in-service platforms, reliable and fast predictions of subsequent motion responses can help to ensure safe operations. Besides, it is also important to predict the motion responses of the designed platforms within a short period to
* Shengwen Xu [email protected] 1
Institute of Urban Smart Transportation and Safety Maintenance, Shenzhen University, 3688 Nanhai Ave, 518000 Shenzhen, China
2
State Key Laboratory of Ocean Engineering, Shanghai Jiao Tong University, 800 Dongchuan Road, 200240 Shanghai, China
3
CIMC Ofshore Co. Ltd, 4008 Menghai Ave, 518000 Shenzhen, China
evaluate the hydrodynamic performances in the early stage of hull design. The underwater part of the semi-submersible platform is mainly composed of columns, pontoons, and transverse braces. A common numerical approach to predict the motion responses is to perform frequency-domain and time-domain analyses based on three-dimensional potential flow theory, supplemented by Morison theory for viscosity correction [1]. However, this method is based on various linearization assumptions and is incompetent for nonlinear problems. In view of this, computational fluid dynamics (CFD) simulations and model tests are usually employed to comprehensively evaluate the hydrodynamic performances of marine structures. However, both CFD simulations and model tests are very t
Data Loading...
