Home > Published Issues > 2020 > Volume 9, No. 1, February 2020 >

Damage Detection Approach for a Mooring Line on an Offshore Structure Using Convolutional Auto-Encoder

Kanghyeok Lee, Minwoong Jung, and Do Hyoung Shi
Department of Civil Engineering, Inha University, Incheon, Korea

Abstract—This study presents a machine learning-based approach to detect damage in mooring lines supporting a floating offshore platform that is installed to collect submarine crude oil. The proposed approach for damage detection using a convolutional auto-encoder can be implemented in three steps: data acquisition, model learning, and model update. The time series data used for damage detection are measured from the environment and the floating offshore platform but not mooring lines due to affordability and efficiency of both installation and maintenance of the sensors on the offshore structure. Therefore, it is expected that the approach proposed in this study can be applied using only data obtained from the structure in an actual environment. 

Index Terms—damage detection, offshore, mooring lines, convolutional auto-encoder

Cite: Kanghyeok Lee, Minwoong Jung, and Do Hyoung Shi, "Damage Detection Approach for a Mooring Line on an Offshore Structure Using Convolutional Auto-Encoder," International Journal of Structural and Civil Engineering Research, Vol. 9, No. 1, pp. 110-113, February 2020. doi: 10.18178/ijscer.9.1.110-113