Extraction of human activity function using approximate inverse filter

摘要

Identification of human activity is important in space control. A model of indoor space was created using a linear system. By applying a combustible gas sensor to this model, we analyzed changes in the indoor environment. The results of this analysis indicate that: (1) the gas diffusion coefficients specific to a given indoor space can be estimated from the response signals of the gas sensor; (2) space transfer characteristics can be described using the gas diffusion coefficient alone; and (3) inverse filtration of the thus estimated space transfer characteristics allows us to extract drive signals from the entire output signals of the sensor. In an experiment in which the indoor environmental changes during daily life were measured, drive signals proportional to the amounts of human activity were obtained. These results suggest that the detection of human actions in an indoor space using a combustible gas sensor can be improved by eliminating the residual components from the entire signal output.