Ambient Lighting Controller Based on Reinforcement Learning Components of Multi-Agents

摘要

Inspired by investigations of thermal comfort, indoor air quality and adequate luminance by using the Predicted Mean Vote Index (PMV) [1-3], the human Ambient Lighting Affect Reward {ALAR) index is proposed for automatic quality control of lighting in the ambient assisted living environment [4, 5]. The ALAR based multi-agent ambient lighting controller is planned also to be used to improve energy savings. Specifically, it predicts the indoor RGB LED lighting conditions at a given time by measuring integrated ALAR index that defines ambient lighting affect to the human. Principles of development of the Ambient Lighting Affect Reward Based Multi-Agent Controller, the ALARBMAC are described in this paper. The ALARBMAC is planned to be applied in the process of development of Eco-social laboratory, the ESLab as a laboratory prototype of the Smart Eco-Social Apartment [4].