A Multi-Agent-Based Intelligent Sensor and Actuator Network Design for Smart House and Home Automation

TLDR

Smart‑house technology seeks to enhance automation, security, and energy efficiency, yet it involves diverse sensors and actuators with conflicting objectives across platforms. The paper proposes a multi‑agent system design framework to enable smart‑house automation. The framework introduces BDI agent behavior models, a regulation‑policy based collaboration mechanism, performance metrics, and is validated through JADE simulations of case studies.

Abstract

The smart-house technology aims to increase home automation and security with reduced energy consumption. A smart house consists of various intelligent sensors and actuators operating on different platforms with conflicting objectives. This paper proposes a multi-agent system (MAS) design framework to achieve smart house automation. The novelties of this work include the developments of (1) belief, desire and intention (BDI) agent behavior models; (2) a regulation policy-based multi-agent collaboration mechanism; and (3) a set of metrics for MAS performance evaluation. Simulations of case studies are performed using the Java Agent Development Environment (JADE) to demonstrate the advantages of the proposed method.