A consensus builder for environmental condition settings in spaces where people with various preferences coexist

2020 Fiscal Year Final Research Report

• Project/Area Number: 19K20257
• Research Category: Grant-in-Aid for Early-Career Scientists
• Allocation Type: Multi-year Fund
• Review Section: Basic Section 60060:Information network-related
• Research Institution: Okayama University
• Principal Investigator: Tarutani Yuya 岡山大学, ヘルスシステム統合科学研究科, 助教 (10751175)
• Project Period (FY): 2019-04-01 – 2021-03-31
• Keywords: Internet of things / 合意形成 / 機械学習

Outline of Final Research Achievements

Various information is collected from IoT devices through the network. As such device becomes more familiar to users, network services are required to consider the influence of user. In this research, we propose a consensus building method　for environmental condition setting in spaces where people with various　preferences coexist.
As a result of our research, we propose a consensus builder using reinforcement learning. We show that our method achieves near-optimal control by controlling air conditioning and lighting. In addition, we show that appropriate control is possible even in situations where complete supervisory data is not available.

Free Research Field

情報ネットワーク

Academic Significance and Societal Importance of the Research Achievements

本研究で取り組んだ合意形成による機器制御は、ユーザの嗜好を直接収集することを必要とせずとも、あらかじめ収集して作成したユーザストレスモデルを用いることで環境内の機器を制御し、消費電力の低減とユーザへの快適な空間の構築を可能とする。また、従来の制御手法では教師あり機械学習を用いた方法を主としていたのに対し、強化学習を用いることでデータの用意が不十分な状況でも適切に制御値を導くことができる手法を提案した。