Prediction of indoor air ventilation performance in kindergarten using nonlinear autoregressive neural network

Abstract

Indoor air pollution has become one of major issues that cause health issues towards the building occupants especially people from the sensitive group such as elderly and younger children. However, indoor air pollutants can be removed by providing adequate ventilation towards the building. Effective and adequate ventilation can help to dilute and removed the pollutants and providing healthier air for the building occupants to breath. Adequacy of ventilation can be determined by measuring the concentration of carbon dioxide (CO₂) in the building as CO₂ is widely used as indicator for ventilation. To determine the ventilation performance, a method of forecasting through modelling process using nonlinear autoregressive neural network (NARNN) is developed where CO₂ concentration data that collected from kindergarten is used construct and find the best fitted model with suitable number neurons and hidden layers. This model can help to predict the future concentration trend of CO₂ in kindergarten and determine the ventilation performance of the building. The concentration of CO₂ in the building is decreasing through the operations hours indicating it has adequate ventilation. The dataset of CO₂ concentration is used to developed a prediction model which consist of artificial neural network (ANN) structure and a model with 1-9-1 structure with data division of 80:20 is the best fitted model for forecasting as it has high accuracy and highly relevant to be used for prediction as it has the nearest R-value near to one. Indoor air quality need a special attention by multiple authorities and organization especially the building that have younger children as occupants. Poor indoor air quality can risk the health of the occupants and disrupt the comfort of occupants on doing their activities in the building. Modelling technique is one of relevant and advance method to forecast the quality of a building as it can help to determine future concentration of pollutants in the indoor environment.