A scalable room occupancy prediction with transferable time series decomposition of CO2 sensor data

Abstract

Human occupancy counting is crucial for both space utilisation and building energy optimisation. In the current article, we present a semi-supervised domain adaptation method for carbon dioxide - Human Occupancy Counter Plus Plus (DA-HOC++), a robust way to estimate the number of people within one room by using data from a carbon dioxide sensor. In our previous work, the proposed Seasonal Decomposition for Human Occupancy Counting (SD-HOC) model can accurately predict the number of individuals when the training and labelled data are adequately available. DA-HOC++ is able to predict the number of occupants with minimal training data: as little as 1 day's data. DA-HOC++ accurately predicts indoor human occupancy for five different rooms across different countries using a model trained from a small room and adapted to other rooms. We evaluate DA-HOC++ with two baseline methods: a support vector regression technique and an SD-HOC model. The results demonstrate that DA-HOC++'s performance on average is better by 10.87% in comparison to SVR and 8.65% in comparison to SD-HOC.