On application of statistically-based mathematical models for power prediction of high-speed craft

This paper presents a comprehensive framework for predicting the power requirements of high-speed craft, including semi-displacement, semi-planing, and planing hull forms, through the use of statistically-based mathematical models. These models belong to the class of reduced-order models, which streamline the inherently complex processes of modeling hull resistance, dynamic trim, and propeller efficiency, including the interactions of multiple design parameters. Model development has relied on various regression techniques and, more recently, artificial neural networks (ANN). These mathematical models often offer advantages over the CFD techniques that nowadays dominate professional journals and therefore deserve careful attention. This review presents the methodology in a structured way to facilitate its integration into power prediction and design-optimization workflows and adopts a user-focused perspective.