A multiagent multisensor based security system for intelligent building

Abstract

The security of buildings, homes, laboratories, offices and factories is essential to human life. An unlucky event is often caused by human negligence. We have developed a multiagent multisensor-based security system or intelligent building. The system can be widely employed in daily life and can detect dangerous situations using sensors. The structure of the security system is divided into four parts, the fire detection/diagnosis agent, intruder detection/diagnosis agent, environment detection/diagnosis agent, and power detection/diagnosis agent. In this paper, we use an adaptive data fusion method in the fire detection/diagnosis agent and use a rule-based method in the intruder detection/diagnosis agent. We use statistical signal detection theory in the environment detection/diagnosis agent, and use a fault detection and isolation procedure (FDIP) in the power detection/diagnosis agent. The security system has a four-variety detection/diagnosis agent. Finally, we implement these methods using computer simulation and achieve quite satisfactory results.