Jayarani Kamble et al Int. Journal of Engineering Research and Applications www.ijera.com ISSN : 2248-9622, Vol. 4, Issue 3( Version 1), March 2014, pp.939-942 www.ijera.com 939 | Page Adaptive Home System Using Wireless Sensor Network And Multi Agent System Jayarani Kamble*, Prof.Nandini Dhole** *(Department of Electronics &Telecommunication RMD Sinhgad School of Engineering, Warje Pune, India) ** (Department of Electronics &Telecommunication RMD Sinhgad School of Engineering, Warje Pune, India) ABSTRACT Smart Home is an emerging technology growing continuously which includes number of new technologies which helps to improve human’s quality of living. This paper proposes an adaptive home system for optimum utilization of power, through Artificial Intelligence and Wireless Sensor network. Artificial Intelligence is a technology for developing adaptive system that can perceive the enviornmrnt, learn from the environment and can make decision using Rule based system.Zigbee is a wireless sensor network used to efficiently deliver solution for an energy management and efficiency for adaptive home. An algorithm used in adaptive home system is based on software agent approach that reduce the energy consumption at home by considering the user’s occupancy, temperature and user’s preferences as input to the system. Keywords – Adaptive Home, Artificial Intelligence, Raspberry Pi, Rule Based System, TARANG Zigbee Module. I. Introduction Energy is consumed every day at home as we perform simple tasks, such as watching television, washing dishes and heating and cooling home spaces during season of extreme weather conditions, using appliances, or turning on lights. Also in wiring system, the equipment will be connected into main power supply directly and many types of wires that people may want to install in wall which increases the cost of a home. Energy saving can be realized by initiating an intelligent system that can detect occupant presence or absence, and adjust the environment accordingly. For example, automatic light switching on or off when a person enters or leaves a room. Customer satisfaction can be achieved by determining the occupant’s comfort level, while customer awareness can be home is a living environment that consists of sensors, actuators, networks and middleware that collect information about the home to provide comfortable living for the occupants. A smart home has the technological capability to adjust itself in certain circumstances to make the home more comfortable for its occupants while sharing a common interface accomplished by real-time monitoring that allows users to see the level and cost of energy they consume on a daily basis. The wireless technology has some amazing achievements in the home automation using ZigBee. The wireless sensor network reduces the cost of the system unit as well as it is much easier to install. The wireless sensor devices have the capability to self- organize into the so-called WSN. Their ability to sense diverse variables of interest, such as the temperature, humidity, pressure, airflow, occupancy, sunlight and other, could greatly improve the limitations of the existing and future energy management systems. Wireless sensor networks are used to monitor psychical parameters such as light and temperature, as well as the presence of users in the home by using infrared sensors in each room. An algorithm is composed of three types of agents. The first type is the personal agent that tracks an individual’s location and learns his/her preferences within the environment. The second type is the local agent that is responsible for a particular zone in the home. The third agent is the central agent that is responsible for decision making and interface with external and internal services. II. Proposed System and Methodology Basically, a Smart Home refers to a home with intelligent to control, monitoring and automate the home system. The Intelligent System proposed is based on the rule-based expert system and unsupervised learning techniques where the problem is how to adapt to new knowledge without destroying the existing knowledge. The core controller unit is equipped with distributed sensors i.e., intelligent agents, which use the rule-based expert system and artificial intelligence concepts to learn and adapt. RESEARCH ARTICLE OPEN ACCESS