Gas-Generating Events Estimation by Means of Remote Sensing and Cepstrum Processing in an Indoor Space

A transfer system composed of fluid is in general a nonlinear field which is described by the Navier-Stokes equations. When, however, the state change in the indoor space is small, it is known that the field can be approximated by a linear field. In this study, the gas transfer system in the indoor space is modeled as a linear system, and an attempt is made at extraction of the gas event generation signal so that the change of the space environment due to daily human activity can be extracted from the observed signal of the inflammable gas sensor. As the first step, it is shown that characteristics of the channel from the indoor gas generation source to the sensor can be extracted by low time windowing of the sensor observation signal through the minimum-phase cepstrum. By applying inverse filtering based on the minimum-phase characteristics of the gas transfer system, the event signal accompanying gas generation can be roughly restored from the observed signal, by reducing the effect of the indoor residual component of the past generated gas. Those properties are verified by a monitoring experiment in an ordinary home. It is expected that the result of this study can be utilized effectively for human activity detection in environmental monitoring using the gas sensor.