Neural Network Model of carbon monoxide distribution in onboard a RO-RO Ship Garage

Abstract

Exhaust emissions in the closed vehicle garage onboard ships not sufficiently ventilated can cause damage to human health and lead even to death and pollution. In this paper, an intelligent artificial neural network model is designed to simulate the carbon monoxide concentration in order to be used to manage the concentration limits in the permissible range. The proposed model is identified based on data collected from real measurements at different load conditions and positions on a Roll-on/Roll-off (RO-RO) vehicle garage onboard a ship vessel. The neural network model is designed, identified and verified to cover all possible operating conditions at different loading and distribution of cars.