Online hydrodynamic forces estimation system based on the artificial lateral line system

Wan，Yuanfang1,2; Jiang，Linfeng1,2; Zuo，Ruyu1,2; Chen，Yongshun1,2; Xie，Guangming1,2,3

2023-11-01

10.1016/j.oceaneng.2023.115677

Ocean Engineering   Impact Factor & Quartile

0029-8018

1873-5258

Vehicles that operate beneath or on the surface of water (e.g., boats, remotely operated vehicles, unmanned underwater vehicles, etc.) are subjected to external hydrodynamic forces and require an online hydrodynamic observer for precise autonomous control. The accurate calculation of current forces requires accurate sensor data and computational fluid dynamics, which impose a huge computational burden and cannot be applied in the real-time control of such vehicles. Fish can accurately perceive the effects of ocean currents, a function that can be achieved through an artificial lateral line system. In this study, an online hydrodynamic observer is proposed that uses an artificial lateral line system to estimate the resultant forces acting on water vehicles in real time for the autonomous control of the vehicle. A small three-layer artificial neural network is used to construct the hydrodynamic observer to evaluate various hydrodynamic forces in the form of a six-axis resultant force, such as the damping effects of currents and effects of wind and waves. In this study, an artificial lateral line system is designed for vehicles of any geometry. The system consists of a multipressure sensor array, data acquisition unit, and computational unit. Each pressure sensor is packaged as a separate unit that can be assembled on the surface of an underwater robot. These pressure sensors measure the pressure changes caused by the ocean current forces, while the data acquisition unit collects, collates, and sends the pressure sensor and attitude and heading reference system (AHRS) data to the processing unit. The processing unit uses an artificial neural network to estimate the ocean current forces acting on the vehicle body coordinate system based on the pressure sensor array, and converts them into the inertial reference system in the geoid based on the data acquired from the AHRS. This study explores the effects of the number of sensors and their locations on the accuracy of the estimated results. It investigates whether a scale factor can be used to estimate other available observers of the same size and geometry to enable the rapid deployment of artificial lateral devices to different vehicles. To evaluate the generalization performance of the proposed hydrodynamic observer, tests were conducted in dynamic water environments, such as rivers and oceans.

Artificial lateral line Artificial neural networks Fast training Generalization performance Online hydrodynamic observer Sensor count optimization

SCI

English

First ; Corresponding

Southern Marine Science and Engineering Guangdong Laboratory (Guangzhou)[KCXFZ20211020174803005];National Natural Science Foundation of China[U1901602];

Engineering ; Oceanography

Engineering, Marine ; Engineering, Civil ; Engineering, Ocean ; Oceanography

WOS:001081783300001

PERGAMON-ELSEVIER SCIENCE LTD

ENGINEERING

2-s2.0-85169779113

Scopus

1.South University of Science and Technology,Shenzhen,1088 Xueyuan Avenue, Guangdong,518055,China
2.Southern Marine Science and Engineering Guangdong Laboratory (Guangzhou),Guangzhou,Guangdong,China
3.Peking University,Beijing,No. 5 Yiheyuan Road, Haidian District,100871,China

Wan，Yuanfang,Jiang，Linfeng,Zuo，Ruyu,et al. Online hydrodynamic forces estimation system based on the artificial lateral line system[J]. Ocean Engineering,2023,287.

Wan，Yuanfang,Jiang，Linfeng,Zuo，Ruyu,Chen，Yongshun,&Xie，Guangming.(2023).Online hydrodynamic forces estimation system based on the artificial lateral line system.Ocean Engineering,287.

Wan，Yuanfang,et al."Online hydrodynamic forces estimation system based on the artificial lateral line system".Ocean Engineering 287(2023).