IEEE System Journal Vol. , No. , April 2012 1 Abstract— Demand response, which is the action voluntarily taken by a consumer to adjust amount or timing of its energy consumption, has an important role in improving energy efficiency. With demand response, we can shift electrical load from peak demand time to other periods based on changes in price signal. At residential level, automated Energy Management System (EMS) have been developed to assist users in responding to price changes in dynamic pricing systems. In this paper, a new intelligent EMS (iEMS) in a smart house is presented. It consists of two parts: fuzzy subsystem and intelligent lookup table. Fuzzy subsystem is based on its fuzzy rules and inputs which produces the proper output for intelligent lookup table. The second part, whose core is a new model of an associative neural network, is able to map inputs to desired outputs. The structure of the associative neural network is presented and discussed. The intelligent lookup table takes three types of inputs which come from fuzzy subsystem, outside sensors and feedback outputs. Whatever is trained in this lookup table are different scenarios in different conditions. This system is able to find the best energy efficiency scenario in different situations. Index Terms— Energy Efficiency, Fuzzy logic, Demand Response, Neural Networks, Smart Grid I. INTRODUCTION MART grid is a novel initiative whose aim is to deliver energy to the users and also to achieve consumption efficiency by means of bidirectional communication [1]. Combination of different hardware devices and software along with an Information and Communication Technology (ICT) infrastructure for a bidirectional communication constitutes the smart grid architecture. ICT has a vital rule in the smart grid architecture as it gives sustainability, creativity and intelligence to it. This electricity network is able to intelligently integrate the actions of all users which are connected to it in order to return them back to users. Users can use this information to optimize their energy consumption. Thus, one of the main objectives of smart grid is encouraging end users to participate in making decision about energy This work was supported, in part, by a grant from CPS Energy through Texas Sustainable Energy Research Institute and Lutcher Brown Chair, the University of Texas, San Antonio, TX, USA. Dariush Shahgoshtasbi is a research assistant in ACE lab at University of Texas at San Antonio, TX 78249 USA (e-mail: isjd@wacong.org). Mo Jamshidi is with the Electrical Engineering Department, University of Texas at San Antonio, TX 78249 USA (e-mail: moj@wacong.org). A preliminary version of this paper appeared at the 6th IEEE international conference on System of Systems Engineering (SOSE), Albuquerque, NM 2011 consumption in an efficient way. But in order to reach energy efficiency, such architecture and interoperability is not enough. We need to add intelligence to it at different levels. At home level, the approach is to add intelligence and then encourage customers to save energy by changing their energy consumption behavior. The electrical grid has two main sides: supply which contains generation, transmission and distribution and demand which consumes the power. The balance between supply and demand sides is necessary at all times, otherwise some blackouts will occur. Throughout the course of a day, when demand increases, the utility companies have to turn on some reserve generation capacity and send the power to the grid to respond to the additional demand. Because these generators usually use gas or diesel to run, they are very expensive. They also emit more CO2 compared to nuclear and hydro power plants, but less CO2 compared to coal-fired power plants used for base load supply. If demand increases and there is not enough capacity, the utilities pay customers to shed load, usually during times of peak load or an emergency situation. Blackouts are the worst case scenario, occurring when demand exceedingly increases and the load cannot be handled. Peak demand traditionally has been a problem in supply-side management, solved by the construction of new power plants. Focusing on management of demand-side can be an alternative way to balance energy use at peak times. So, demand response can be in response to an economical signal which is mostly a pricing signal. By using demand response, we are able to shift electrical load from peak demand time to other periods which reduces the ratio of peak to average load. This can be resulted in improving efficiency, reducing costs [2] and risk of outages. Demand response can be done at different levels like generation, transmission or end user level. A lot of work has been done at generation and transmission levels [1,3-7]. At the end user level, the smart grid is not only able to provide information about electricity consumption for both users and network operators, but also can dispatch renewable energy resources to the grid. At the residential or end user level demand response, challenges that should be considered include real time pricing information to consumer, networking home devices, security and implementing automated EMS. In this paper, a new intelligent EMS system, called iEMS is presented. It takes inputs from the grid and by using an intelligent algorithm, tries to find effective and efficient energy consumption. It can also match users’ preferences and behaviors and then find optimal energy scheduling according to the dynamic price. This approach is useful especially in a dynamic pricing system which modification of energy consumption is unrecognized by a A new intelligent Neuro-Fuzzy paradigm for Energy Efficient Homes Dariush Shahgoshtasbi, Student Member, Mo Jamshidi, Fellow, IEEE S