A New Ridgely-Nevitt Regression-Based Computational Tool for Resistance and Power Predictions for Trawlers
Keywords:Ridgely-Nevitt regression, Computational procedure, Fishing vessels series, Resistance, Hull model
This paper aims to present the architecture and the prediction accuracy of a new computational procedure of the “Ship Power V 1.0” software based on the Ridgely–Nevitt regression, applied to hull resistance predictions for Ridgely-Nevitt series trawlers. The Ridgely-Nevitt series is an important series of trawlers developed by and tested at the Webb Institute. Experimental resistance data have also been presented in the form of a regression model used to develop a new procedure of the “Ship Power V 1.0” software. Furthermore, this new procedure was completely harmonized with other software procedures based on the Holtrop and Van Oortmerssen evaluation methods. Although the mathematical formulation of the Ridgely-Nevitt regression model allows the assessment of the residual resistance coefficient of only nine values from the speed-length ratio, the implementation of an interpolation procedure made possible resistance predictions for any speeds from the acceptable speed-length ratio range. Resistance prediction accuracy improvement introduced by the new procedure was proven by the validation of calculation results not only against experimental data but also against the prediction results of other software procedures for three hulls from the Ridgely-Nevitt fishing vessel series. MAPE (mean absolute percentage error) values calculated against experimental data for the analyzed models were 3.26, 1.71, and 3.36, respectively. Prediction result comparisons of the Ridgely-Nevitt regression-based “Ship Power V 1.0” computational procedure and the experimental data and predictions of other computational procedures performed on three hulls from the Ridgely-Nevitt series have shown a substantial improvement in prediction accuracy.
How to Cite
Copyright (c) 2023 Transactions on Maritime Science
This work is licensed under a Creative Commons Attribution 4.0 International License.