音響周波数応答から室温を推定する手法を開発した林田君（東工大 OB）の論文が Acoustical Science and Technology に掲載されました。
A. Hayashida, Y. Mizuno, and K. Nakamura, “Estimation of room temperature based on acoustic frequency response,” Acoust. Sci. Technol., vol. 41, no. 4, pp. 693-696 (2020).
Monitoring various types of ‘states’ of rooms is highly required for a number of reasons. First, detecting the state of persons is one of the most important applications. The activity or health condition of persons in a room has become a major issue in aging societies. The use of surveillance cameras is not acceptable for privacy protection. Acoustic surveillance has been studied as an alternative method, where the sounds generated by persons’ activities are passively detected and analyzed using sophisticated signal processing. Second, the physical states of a room itself need to be monitored for purposes of security and air conditioning. It is useful to remotely detect whether the windows or doors of a room are open or closed. To achieve this, every window/door needs to be equipped with a sensor. The temperature of the room is also essential information for air conditioning and for detecting fire or excessive heat. Methods of estimating air temperature using sound propagation time have already been utilized. However, the method is not suitable for grasping the spatial temperature distribution. In this study, we attempt to measure temperature and its distribution in a room from acoustic frequency responses. The point of this study is that it focuses on the modal responses of the room instead of directive detection using microphone arrays. The system is composed of one loudspeaker and one microphone, which obtain the acoustic frequency responses of the room; they will be replaced by a smart speaker or a similar device in the future. First, the measurement principle is explained. Second, the method is demonstrated using a 1/8-scale model. Finally, a trial for a practical room is performed.