Using Remote Monitoring and Machine Learning to Classify Slam Events of Wave Piercing Catamarans

Abstract

An onboard monitoring system can measure features such as stress cycles counts and provide warnings due to slamming. Considering current technology trends there is the opportunity of incorporating machine learning methods into monitoring systems. A hull monitoring system has been developed and installed on a 111 m wave piercing catamaran (Hull 091) to remotely monitor the ship kinematics and hull structural responses. Parallel to that, an existing dataset of a similar vessel (Hull 061) was analysed using unsupervised and supervised learning models; these were found to be beneficial for the classification of bow entry events according to key kinematic parameters. A comparison of different algorithms including linear support vector machines, naïve Bayes and decision tree for the bow entry classification were conducted. In addition, using empirical probability distributions, the likelihood of wet-deck slamming was estimated given a vertical bow acceleration threshold of 1 in head seas, clustering the feature space with the approximate probabilities of 0.001, 0.030 and 0.25.

Bow Slamming Detection and Classification by Machine Learning Approach

“…In Shabani et al (2021), from measurements on-board a wave-piercing catamaran six variables were selected for the classification of the water entry of the center-bow. Here, a similar set is considered and further extended to include information about ship's rigid-body motion and speed, as well as typical parameters relative to stochastic sea-state.…”

“…The use of Machine Learning methodology offers the advantage to search for different types of slamming events by using the same plain and robust data preparation pipeline. Shabani et al (2021) applied unsupervised ML to identify distinct classes of relative motion events from monitoring data collected on a wave piercing catamaran during full-scale trials in the Canary Islands. Thus, events labeled according to this classification were employed to train the algorithm in a supervised learning approach, showing good performance in recognizing those events on the test dataset.…”