Optimized Integration of Electric Vehicles into Smart Grid

Electric Vehicles (EVs) are considered as a promising solution for the current fossil-based energy shortage and pollutant emission problems. Batteries in EVs not only can support daily commute of EV customers, but can also provide flexible charging rates and storage capabilities that can be effectively harnessed by the power grid when EVs are integrated into the grid. To maximize the benefits of using EVs, regulated and optimized charging control needs to be provided by load aggregators for connected vehicles. An EV charging network is a typical cyber-physical system, which includes a power grid and a large number of EVs as well as aggregators that collect information and control the charging procedure. The Information and Communication Technologies (ICT) that are currently under development for future smart grid systems can enable load aggregators to have bidirectional communications with both the grid and EVs to obtain real-time price and load information, and to adjust EV charging schedules in real time. A framework based on EV aggregation for implementing vehicle-to-grid (V2G) integration is presented in the dissertation. The framework supports the effectiveness of EV aggregation while keeping the information of each EV unrevealed to the grid. The characteristics of EV battery and energy market are analyzed to provide the basis for determining which services that EVs can provide to the grid. We deploy a probabilistic modeling approach to take into consideration the variability associated with the behavior of the EVs. In this dissertation, we comprehensively investigate EV charging scheduling prob- lems for the provision of various scenarios: EV supporting system frequency regulation while satisfying customers\u27 demand, EV trading in multi-settlement energy market, EV improving renewable energy efficiency and EV smoothing renewable power output. We set the objective as to maximize the benefit regarding EV integration to the grid by jointly considering the EV battery constraints, customers\u27 demands and system constraints. In addition, we present a communication protocol for interactions among different entities regarding EV integration, and demonstrate how to integrate the proposed scheduling approach in real-time charging operations. To achieve our goals, we formulate our schemes based on the following problem solving techniques: Linear Programming (LP), Mixed Integer Linear Programming (MILP), heuristic algorithm based LP rounding, Lyapunov optimization and Model Predictive Control (MPC) to provide optimal solutions. Extensive simulation results based on real data are presented to justify the effectiveness of the proposed approaches and to show how several key parameters affect their performance.A framework based on EV aggregation for implementing vehicle-to-grid (V2G) integration is presented in the dissertation. The framework supports the effectiveness of EV aggregation while keeping the information of each EV unrevealed to the grid. The characteristics of EV battery and energy market are analyzed to provide the basis for determining which services that EVs can provide to the grid. We deploy a probabilistic modeling approach to take into consideration the variability associated with the behavior of the EVs. In this dissertation, we comprehensively investigate EV charging scheduling problems for the provision of various scenarios: EV supporting system frequency regulation while satisfying customers\u27 demand, EV trading in multi-settlement energy market, EV improving renewable energy efficiency and EV smoothing renewable power out- put. We set the objective as to maximize the benefit regarding EV integration to the grid by jointly considering the EV battery constraints, customers\u27 demands and system constraints. In addition, we present a communication protocol for interactions among different entities regarding EV integration, and demonstrate how to integrate the proposed scheduling approach in real-time charging operations. To achieve our goals, we formulate our schemes based on the following problem solving techniques: Linear Programming (LP), Mixed Integer Linear Programming (MILP), heuristic algorithm based LP rounding, Lyapunov optimization and Model Predictive Control (MPC) to provide optimal solutions. Extensive simulation results based on real data are presented to justify the effectiveness of the proposed approaches and to show how several key parameters affect their performance