Early detection of parametric roll by application of the incremental real-time Hilbert–Huang Transform. (1st February 2016)

• Record Type: Journal Article
• Authors: Yu, Liwei; Ma, Ning; Gu, Xiechong
• Is Part Of: Ocean engineering. Volume 113(2016)
• Page Start: 224
• Page End: 236
• Publication Date: 2016-02-01
• Subjects: Parametric roll -- Time-frequency analysis techniques -- Real-time Hilbert–Huang transform -- Unified Model -- Early detection -- Advance warning; Ocean engineering -- Periodicals; Ocean engineering; Periodicals; 620.4162
• DOI: 10.1016/j.oceaneng.2015.12.050
• Languages: English
• ISSNs: 0029-8018

Abstract

Abstract: This work focuses on the early detection and advance warning of parametric roll in regular and irregular head seas. The time–frequency dependent analysis technique, Hilbert–Huang Transform (HHT) is applied to acquire the so-called "instantaneous frequency", which can be used for the detection of the non-linear parametric roll motion. Firstly, an incremental real-time HHT (IR-HHT) technique is implemented to an artificial time series which is simply composed from the parametric roll and pitch motion time series with its main characteristics: frequency shift and amplitude growth. This is followed by the proposal of a detection scheme aimed at the identification of the frequency shift and amplitude growth between roll and pitch motion. The proposed detection scheme is validated by analyzing parametric roll time series obtained from numerical simulation in regular and irregular head seas. Results show that the onset of parametric roll can be detected successfully through the detection scheme based on IR-HHT technique. Moreover the robustness of the detection scheme is validated. Highlights: A detection scheme for parametric roll using Hilbert–Huang Transform is proposed. The scheme is designed to detect the frequency discontinuity between pitch and roll. Validations using numerical model show parametric roll is detected at onset stage. The robustness limits of the factors in the detection scheme are discussed. Robustness of the scheme is validated by applying it to conventional roll motion.