1551-3203 (c) 2017 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information. This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TII.2017.2755465, IEEE Transactions on Industrial Informatics IEEE TRANSACTION ON INDUSTRIAL INFORMATICS: TII-17-0125 1 Demand-Side Management by Regulating Charging and Discharging of the EV, ESS, and Utilizing Renewable Energy Mosaddek Hossain Kamal Tushar, Member, IEEE, Adel W. Zeineddine and Chadi Assi, Senior Member, IEEE Abstract—The evolution in microgrid technologies as well as the integration of electric vehicles (EVs), energy storage systems (ESSs) and renewable energy sources (RESs) will all play a significant role in balancing the planned generation of electricity and its real-time use. We propose a real-time decentralized demand-side management (RDCDSM) to adjust the real-time residential load to follow a pre-planned day-ahead energy gener- ation by the microgrid, based on predicted customers’ aggregate load. A deviation from the predicted demand at the time of consumption is assumed to result in additional cost or penalty inflicted on the deviated customers. To develop our system, we formulate a game with mixed strategy which in the first phase (i.e., prediction phase) allows each customer to process the day ahead raw predicted demand to reduce the anticipated electricity cost by generating a flattened curve for its forecasted future demand. Then, in the second stage (i.e., allocation phase), customers play another game with mixed strategy to mitigate the deviation between the instantaneous real-time consumption and the day-ahead predicted one. To achieve this, customers exploit renewable energy and energy storage systems and decide optimal strategies for their charging/discharging, taking into account their operational constraints. RDCDSM will help the microgrid operator better deals with uncertainties in the system through better planning its day-ahead electricity generation and purchase, thus increasing the quality of power delivery to the customer. We evaluate the performance of our method against a centralized allocation and an existing decentralized EV charge control non-cooperative game method both which relies on a day ahead demand prediction without any refinement. We run simulations with various microgrid configurations, by varying the load and generated power and compare the outcomes. Index Terms—Microgrids, Smart grids, Electric Vehicles, Game theory, Optimization, Renewable energy sources, Energy storage, Home Energy Management System, Demand-Side man- agement, Demand forecasting, Mixed strategy. I. I NTRODUCTION T He rapid surge in demand for electricity is considered as one of the most significant problems that are facing the power grid. To achieve higher reliability, robustness, and stability today’s power grids are designed to serve peak de- mands rather than the average load. This can result in a power generation and distribution system that is underutilized as well as in the waste of natural resources [1], [2]. Hence, utility com- panies are continuously adjusting the power generation of their Manuscript received January 23, 2017; revised June 16, 2017; accepted September 02, 2017. This work was supported in part by the Natural Sciences and Engineering Research Council (NSERC) of Canada Discovery Grant and in part by Concordia University. M. H. K. Tushar, A. W. Zeineddine and C. Assi are with Concordia University, Montreal, QC H3G 2W1, Canada plants to balance the total loads and their variations. Indeed, fast-responding generators such as fossil-fuel generators are costly and have a significant greenhouse gas (GHG) footprint [3]. Power system planners are therefore facing a pressing challenge to meet their customers surging demands while ensuring electricity systems integrity. Numerous methods have been proposed to alleviate problems of uncertainties in power system and regulate users’ consumption profiles. The aim of these plans (such as demand-side management or DSM) is to deploy the current capacity more efficiently without modifying the existing grid infrastructure [4], [5], [6], [7], [2]. The evolution of the smart grid, integration of RES, AMI smart meters, EVs and dynamic pricing, have all added momentum to solve the DSM problem efficiently. Further, the wide spread deployment of home energy management systems (HEMS) and communicating devices will upgrade the existing power grid structure and transform it into a more intelligent and decentralized system [8]. Recently, new distributed entities that have not existed pre- viously, e.g., microgrids and active distribution networks, are becoming essential components of the smart grid. A microgrid is a miniature form of the smart grid which enables a two-way communication to exchange control and information between customers and the operator. The integration of RESs, ESSs, and EVs as key components in the microgrid, can lead to imbalances in the system (e.g., due to the stochastic nature of RES, randomness arising from EV’s behavior, etc.) and further aggravates the problem of load uncertainties. ESSs and EVs can however also present new opportunities for the DSM and can be employed to store energy when demand is low compared to the amount of production from RESs and discharge it in times of shortage or peak demands [9]. Now, the amount of renewable energy, electricity demand for residential use, charging and discharging of EVs, all vary at the time of operation compared to the day-ahead prediction [10], [11], [9]. Typically, the energy must be either consumed or stored at its precise production moment, whereas the demand for it varies throughout the day and across the seasons [9], [10]. As a result, the utility-customer interactions vary depending on the timescales, as well as the types of customers units [12]. Most efficient forecast methods do not accurately predict the RES generation, household, and EV demands; therefore, a discrepancy exists between the predicted load and the actual use of electricity by the customers of the grid[13]. Moreover, the variation between predicted and actual load is primarily dependent on the forecast methods and quality. This can be Copyright c 2009 IEEE. Personal use of this material is permitted. However, permission to use this material for any other purposes must be obtained from the IEEE by sending a request to pubs-permissions@ieee.org.