Recently, human-centered technology development has been in the limelight. It reflects user characteristics as much as possible and considers physical and psychological characteristics of humans even their current emotional state. For example, there is a sleeping room service that adjusts the temperature, humidity, and lighting according to the user’s biorhythms, as well as a, light system that applies lighting therapy technology to an apartment. This paper was based on existing research, which suggests that light and sound conditions affect the emotional state of human subjects. The EEG of biological signals is taken, and the individual biological properties are analyzed using concentration, depression, and relaxation indexes. According to the results, a system to automatically control light and sound and select music is proposed. After analyzing the state of the user by measuring the EEG, the system adjusts the living environment according to the result. First, the index is calculated by collecting the normal EEG of the user. EEG analysis is sorted by frequency component data in accordance with the fast Fourier transform method to grasp the density and distribution of the components. The concentration, depression, and relaxation indexes are calculated using this information. The error range based on the average of each index is decided, making it possible to analyze the status of the measured EEG in real time. The system optimizes the illumination and volume in the living environment for the four types of statuses - concentration, depression, relaxation, and normal -and allows the selection of music for the user’s status. The songs are played at random among the songs in four status groups: A (concentration), B (relaxation), C (depression), and D (normal). An effective song for each status is chosen automatically from the appropriate song list. In the following figure, the right table shows proper illumination and volume that feels comfortable for indoor tasks and the left picture is the algorithm of the control system designed using the table. Fig. 1. System Algorithm and Lighting and Sound proper size The EEG measuring device used in the experiment is Emotiv’s EPOC. It is composed of 14 channels and two ground sensors, and receives data at the rate of 128Hz/s. We use EEG measured from the F3, and F4 channels based on the existing theory that the EEG is closely related to the frontal lobe, and thus is associated with concentration and depression. The experiment was conducted over three days on a female in her 20s. We applied the developed system on the third day according to the measured EEG, using the EEG measured over the previous two days as the baseline. In this paper, we describe the development of a living environment automatic control system that analyzes user status through the EEG of biological signals and intelligently controls the lighting and music volume and selection. Such a system could be utilized in a variety of fields. For example, an intelligent apartment could provide a comfortable indoor environment as well as save energy, light and sound therapy can be used to treat depression and somnipathy. However, the accuracy of classifying mental state based on the measured EEG is low. To increase the reliability of such EEG analysis and more accurately classify the status of the user, we should apply a nonlinear algorithm such as SVM or HMM according to probability. Currently, we are studying how to apply the HMM algorithm.

* ITRC  기여율 = 1