0018-9545 (c) 2016 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/TVT.2016.2636874, IEEE Transactions on Vehicular Technology 1 HELOS: Heterogeneous Load Scheduling for Electric Vehicle-integrated Microgrids Gaoxiang Li, Di Wu, Member, IEEE, Jiefeng Hu, Senior Member, IEEE, Yong Li, Member, IEEE, M. Shamim Hossain, Senior Member, IEEE, Ahmed Ghoneim Member, IEEE, Abstract—With the increasing concerns on the worldwide envi- ronmental conditions and rapid development of renewable energy technologies, microgrids have been regarded as a promising solution to reduce the burden of infrastructure-based power systems. However, due to the intrinsically intermittent features of existing renewable energy, along with random residential be- havior patterns, unpredictable plugged-in or unplugged actions of Electric Vehicles (EVs) and time-varying electricity price, it is challenging for microgrid operators to efficiently perform load scheduling and energy management. In this paper, we propose an online algorithm to conduct cost-aware scheduling of EV loads and energy supplies for microgrids. We formulate this problem into a stochastic optimization problem with the objective of minimizing the time-average cost of a microgrid, including the electricity purchase cost from the main grid, charging and discharging cost of batteries, renewable harvesting cost and life- cycle greenhouse-gas emission cost. To solve this problem, the key idea is to exploit the dynamics of electricity price to conduct the battery charging and discharging operations, renewable energy harvesting and schedule EV loads properly. Our method is based on the Lyapunov optimization technique, which has low computational complexity and only requires limited prediction of price information. The theoretical analysis of our algorithm confirms that the proposed strategy can achieve the optimality with explicit bound. By conducting extensive real-data driven simulations, we demonstrate that our proposed algorithm can achieve much lower cost and be more eco-friendly than other alternative solutions. Manuscript received November 12, 2015; revised May 18, 2016 and October 03, 2016; accepted November 28, 2016. This work was supported in part by the National Key Research and Development Program of China under Grant 2016YFB0201900, the National Science Foundation of China under Grant 61272397, Grant 61572538. The associate editor coordinating the review of this paper and approving it for publication was M. Benbouzid. (Corresponding author: Di Wu.) G. Li and D. Wu are with the Department of Computer Science, Sun Yat-sen University, Guangzhou 510006, China, Guangdong Province Key Laboratory of Big Data Analysis and Processing, Sun Yat-sen University, Guangzhou 510006, China, and also with Collaborative Innovation Center of High Performance Computing, National University of Defense Tech- nology, Changsha 410073, China. (e-mail: ligx3@mail2.sysu.edu.cn, wu- di27@mail.sysu.edu.cn). J. Hu is with the Department of Electrical Engineering, The Hong Kong Polytechnic University, Hung Hom, Hong Kong (e-mail: jer- ry.hu@polyu.edu.hk). Y. Li is with the Department of Electronic Engineering, Tsinghua Univer- sity, Beijing, China (e-mail: liyong07@tsinghua.edu.cn). M. S. Hossain is with the Department of Software Engineering, College of Computer and Information Sciences (CCIS), King Saud University, Riyadh 11543, Saudi Arabia (e-mail: mshossain@ksu.edu.sa). A. Ghoneim is with the Department of Software Engineering (SWE), College of Computer and Information Sciences (CCIS), King Saud Univer- sity, Riyadh 11543, Saudi Arabia and with Computer Science Department, College of Science, Menoufia University, Menoufia 32721, Egypt (e-mail: ghoneim@ksu.edu.sa). Copyright (c) 2015 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 an email to pubs-permissions@ieee.org. Index Terms—Electric vehicle, load scheduling, energy man- agement, microgrid I. I NTRODUCTION In the past decade, much attention has been paid to dis- tributed power generations. Meanwhile, due to the stressed power network and technology development, the penetration of renewable energy resources is increasing sharply. According to the world energy outlook [1], wind and solar will provide 25% and 7.5% of the total power generated by renewables in 2035 respectively. In such a context, as a new form of smart grid, Microgrid has been proposed with the aims of achieving high energy efficiency and low emission. A microgrid integrates the local renewable energy generation, traditional energy generation (e.g., diesel generators), energy storage devices and residential load into a single controllable unit that can be operated in either grid-connected mode or isolated mode [2]. Microgrids can be divided into three categories, namely, remotely-located microgrid, industrial microgrid and utility microgrid [3]. The remotely-located microgrids only operate in the isolated mode since they are far from the main grid, and the transmission loss will cause lots of energy wastage. Therefore, a remotely-located microgrid usually generates energy by local generators. The industrial microgrids need to serve critical and delay-sensitive loads [3], and use locally generated energy to provide reliable energy supply. A utility microgrid can be viewed as one part of a distributed grid system and provide local support to the main grid during electricity peak hours. Thus it can operate under both modes. A utility microgrid can draw energy from the main grid and usually be equipped with renewable energy generation and energy storage devices, which are more eco-friendly than diesel generators [4]. In this paper, the utility microgrids are focused. As shown in Fig.1, the architecture of a utility microgrid includes re- newable energy generation, energy storage devices, Central Management Controller (CMC), power lines to the main grid and residential houses and EVs. In the connected mode, if the local energy supply is insufficient, the CMC will draw energy from the main grid to meet the residential and EV load requirements. On the other hand, if the generated energy in the microgrid is abundant, the excessive energy can be sold back to the main grid [4]. From the perspective of a microgrid operator, the total cost of a microgrid results from the following four sources: purchasing electricity from the main grid, charging and dis- charging batteries, harvesting renewable energy, and life-cycle