An ANN-Based Model to Artificially Transform a Floating Vessel into a Wave Monitoring Buoy

Abstract

Floating platform offshore operations, such as load transfer from a supply boat and hydrocarbon offloading operations, are often limited by sea conditions. Access to sea condition information from a nearby wave/environmental monitoring buoy is not always available or there may be delays in data transmission. In the absence of reliable and accurate data from a nearby wave buoy, the motions of a floating platform/vessel may be used to estimate the sea conditions. This paper presents a method in using Artificial Neural Network (ANN) models to map the motions of a floating vessel to the wave elevations and to eventually estimate the significant wave heights of the sea the floating vessel is in. The ANN models can be trained using either: (1) measured data in the form of measured vessel motions and data from a nearby wave buoy, or (2) simulated data, i.e. vessel motions computed using numerical simulations. In this paper, demonstration of the ANN method uses simulated data under various sea conditions. The trained ANN models are tested for sea conditions that are not part of the training data, and the ANN predictions are found to be very close to the results calculated using numerical simulations. This is an important step to show that the trained ANN models have learned the presented information and can generalize the knowledge. The methodology presented in this paper may be used to establish an ANN model for estimation of the significant wave heights based on the measured motions of a floating vessel.