Slamming classification dataset
Description
The file contains 11 (eleven) columns, each one related to the features extracted from the analysis of the time series directly or indirectly measured on a ship scaled physical model tested in the wave basin (towing tank) at the Institute of Marine Engineering of the Consiglio Nazionale delle Ricerche (National Research Council), Rome, Italy (formerly, INSEAN). The dataset allows the user to set up a procedure for the identification of slamming events by using Machine Learning algorithms, as explained in the paper entitled "Bow slamming detection and classification by Machine Learning approach" by Dessi, Sanchez-Alayo, Shabani and Lavroff in Ocean Engineering, 2023, 287, 115646. The reader can also refer to Table 5 for further explanation of the columns. The last column (12) indicates what type of slamming event occurred, and is necessary to employ a supervised learning approach, as described in the paper. All the data are expressed at the scale of the physical model tested in the towing-tank. The tests were carried out in 2002 but Machine Learning analysis has been done recently. The dataset is a collection of several towing-tank tests relative to different speed and sea state conditions as reported in Table 4 of the cited paper. Fewer significant digits for numbers in certain columns indicate the suggested maximum precision (VAR09-10-11). The column relative to the frequency at which the wavelet peak was calculated (VAR06) reports repeated values of the (peak) frequencies. This depends on the frequency quantization of the wavelet algorithm.
Files
Steps to reproduce
The raw data were obtained from an experimental campaign performed at the CNR-INM (former INSEAN) wave basin on a scaled elastic model. Information about the scaled model have been published through the years and relevant papers are mentioned in the paper (OE 2023) which the published data are directly related to. Test conditions are also reported in the paper. The published data were obtained after feature extraction according to the pipeline illustrated in the paper.