Zhiqiang Hu
Marine, Offshore and Subsea Technology, School of Engineering, Newcastle University, UK
During year of 2020, although the pandemic brought some negative influences on the international collaborations, but the collaborative works between RIAM Kyushu University and Newcastle University have been carrying out steadily. The outcomes are listed as below.
1) Attendance of The 2nd International Symposium on Novel Computational and Experimental Methods for Complicated Fluid-Structure Interactions, at RIAM Kyushu University in Jan 2021 and was held online.
2) During the collaborative research in year 2020, a new AI technology-based method, named SADA, was proposed for the prediction of dynamic responses of FOWTs. Currently, it is SADA version 1.0. The methodology of SADA starts with the selection of Key Disciplinary Parameters (KDPs). The AI module in SADA was built by a coupled aero-hydro-servo-elastic in-house program DARwind with the machine learning algorithms deep deterministic policy gradient (DDPG). This latest research outcome has been presented in the RIAM conference in Jan 2021.
3) Based on the research in year 2020, a collaborative conference paper entitled ‘Software-in-the-Loop combined Machine learning for Dynamic Responses Analysis of Floating Offshore Wind Turbines’ OMAE2021-65524 has been submitted to OMAE Conference 2021. Prof Changhong Hu is the co-author of this paper. At this moment, the paper is under the revision status. If it can be accepted, it will be presented by Mr. Peng Chen, a PhD student of Zhiqiang Hu at Newcastle University in June 2021.
1. RIAM international symposium
The 2nd International Symposium on Novel Computational and Experimental Methods for Complicated Fluid-Structure Interactions was successfully held online on 29th Jan 2021.
Professor Zhiqiang Hu attended this symposium and gave an presentation with topic on
‘Software-in-the-Loop combined Reinforcement Learning Method for Dynamic Response Analysis of Floating Wind Turbines’. This presentation introduced the collaborative research outcomes of Zhiqiang Hu’s research team at Newcastle University and Professor Changhong Hu’s research team at Kyushu University during year 2020. This presentation introduced the in-depth knowledge, structure of SADA and challenges need to be overcome in the future. After the presentation, Zhiqiang Hu made fruitful discussions with other scholars online. After the symposium, professor Changhong Hu and Zhiqiang also made a detailed discussion and confirmed to continue the collaborative research in the field of AI-based knowledge and floating wind turbines.
2. Collaborative research on AI technology for floating wind turbine
In the year of 2020, a collaborative research on AI-based technology for dynamic responses analysis of floating wind turbine was conducted very well and fruitful research outcomes have been achieved. The SADA method is newly proposed for the prediction of dynamic responses of FOWTs with a smart application of artificial intelligence technology. Another new skill, the selection of Key Disciplinary Parameters (KDPs) is also proposed as the first key skill in SADA.
The AI module in SADA was built by a coupled aero-hydro-servo-elastic in-house program DARwind with the machine learning algorithms deep deterministic policy gradient (DDPG).
Basin experimental data was used in SADA for the AI training. The SADA method was used to analyze the dynamic responses of Hywind 5MW wind turbine and the results showed very satisfactory.
In addition, thanks for the support and collaboration in 2020, a collaborative conference paper submitted to the 40th International Conference on Ocean, Offshore and Arctic Engineering (OMAE2021). The paper is titled ‘Software-in-the-Loop combined Machine learning for Dynamic Responses Analysis of Floating Offshore Wind Turbines’. This paper is under revision status. If it can be accepted for publication, the first author will give a presentation online in the OMAE2021 conference.
In year 2021, the collaborative research will be conducted to a further step. Currently, there are still some challenges in the anticipation of using SADA method for a higher accuracy. Some investigations on statistical criterion application in error estimation will be conducted as a priority. The aerodynamic influence will also be investigated to a further detailed extend. The AI technology application in offshore renewable energy is also the strong side of Professor Changhong Hu’s research team. So, it is expected the novel method currently proposed will have a potential opportunity to be used in some other offshore renewable energy facilities, such as ocean wave energy converter.
国際化推進共同研究概要
No.21 20RE-9
タ イ ト ル： Novel computational and experimental approaches to the analysis of complicated fluid-structure interactions
研究代表者： WAN Decheng
所内世話人： 胡 長洪
研究概要：
本国際化推進共同研究について今年度は 2 年目で，共同研究・研究集会とも予定通り実 施した。共同研究について、海洋工学分野での流体・構造連成解析問題に関する CFD 手 法開発に関する検討が行われ、関連の研究成果は４編の国際学会論文に纏められ採択さ れた。R3 年 1 月 29 日にオンライン国際研究集会「The 2nd International Symposium on Novel Computational and Experimental Methods for Complicated Fluid-Structure Interactions」が開催され、外国から 22 名、日本から 25 名の参加者があり、流体・構 造連成解析問題に関する数値解析方法及び水槽実験方法の高精度化・高効率化に関して 有意義な国際研究集会となった。