AI-based Building Management and Information System with Multi-agent Topology for an Energy-efficient Building: Towards Occupants Comfort

Abstract

Most of the electrical energy consumption occurs in residential buildings in maintaining the desired comfort level for occupants. Since comfort level and energy consumption are conflicting in nature, there is a need for energy-efficient building management and information system. In this article, Artificial Intelligence (AI) based building management and information system with multi-agent topology for the energy-efficient building is proposed. The multi-agent topology building management and information system are based on minimizing the energy consumption and maximizing the level of comfort by reducing the error between the actual parameters of the environment and the desired environmental parameters. Firstly, the constrained nonlinear optimization algorithm is applied in the first optimization, and secondly the optimization using artificial intelligence incorporating deep learning concept training and validation to obtain a set of optimized solutions. These solutions comprise values of temperature, illumination level, and concentration of CO2 for maximum comfort level in terms of thermal, visual and air quality and minimum energy consumption at the same time. The developed system is energy efficient and maintains a high comfort level for occupants.