Prediction of Dynamic Responses of Flow-Induced Vibration Using Deep Learning

Gi-Yong Kim,Eun Soo Kim,Chaeog Lim,Sung-Chul Shin

doi:10.3390/app11157163

Abstract

Flow-induced vibration (FIV) is a phenomenon in which the flow passing through a structure exerts periodic forces on the structure. Most studies on FIVs focus on suppressing this phenomenon. However, the Marine Renewable Energy Laboratory (MRELab) at the University of Michigan, USA, has developed a technology called the vortex-induced vibration for aquatic clean energy (VIVACE) converters that reinforces FIV and converts the energy in tidal currents to electrical energy. This study introduces the experimental data of the VIVACE converter and the associated method using deep neural networks (DNNs) to predict the dynamic responses of the converter. The DNN was trained and verified with experimental data from the MRELab, and the findings show that the amplitudes and frequencies of a single cylinder in the FIV predicted by the DNN under various test conditions were in good agreement with the experimental data. Finally, based on both the predicted and experimental data, the optimal power envelope of the VIVACE converter was generated as a function of the flow speed. The predictions using DNNs are expected to be more accurate as they can be trained with more experimental data in the future and will help to substantially reduce the number of experiments on FIVs.

Highlights

Interest in marine renewable energy has gradually been increasing worldwide, and various marine technologies have been developed for commercialization of the energy generated thereof
This study involved the development of a deep learning model that can predict the cylinder amplitudes and frequencies by training the flow-induced vibration (FIV) experimental data from the Marine Renewable Energy Laboratory (MRELab) team with a neural network
Using four input values for deep learning model training, we performed data preprocessing, and the cylinder amplitudes and frequencies were predicted as output values

Summary

Introduction

Interest in marine renewable energy has gradually been increasing worldwide, and various marine technologies have been developed for commercialization of the energy generated thereof. Through tefns,twiantegr under the above c0o.8n9d3itions,1t.h0e93parame1t.e2r2s7to train1.t4h1e2tested1d.5a4ta of the single-cylinder amplitude and natural frequency of the system for the deep learning modeTlhwroeureghfirtsetsteixntgraucntedde.r Athetoatbaloovfe scioxnpdaitriaomnse,tethrse wpaerraemexetterarsctteodt,raasinfothlleowtess:tesdprdinatga coofnthsteasnitngKl,eh-cayrlninedsserdaammppliintugdreaatinodonf atthueraslyfsrteeqmueζnhcaryneosfs ,thneusmysbteermofforrotthaetidoneespolfetahreniinngduction generator to circulate the water in the LTFSW channel fmotor, natural frequency of the cylinder in water, fn,water, reduced velocity U*, and the Reynolds number Re. Correlation analysis was performed to identify the effects of the extracted parameters on the dependent variables, amplitude (A/D), and frequency (fcyl) of the cylinder. After verifying the general performance of the trained network, we input the untested data to predict the amplitude and frequency of the cylinder.

Deep Learning Model Structure

Deep Learning Model Training

Findings

Conclusions