Behzad Heydaryan Manesh

• This thesis proposes an economically optimal generic energy management framework to enable industrial, commercial, and residential producers/consumers (prosumers) to offer their power and energy flexibility in different energy markets, with a particular focus on mid-term and real-time energy markets and two main use-cases: beverage production plants and bidirectional electric vehicle (EV) charging hubs. The proposed framework utilizes the theory of model predictive control (MPC) in a hierarchical structure with different prediction horizons, which enables long-term and mid-term energy planning as well as real-time operational control. The optimal control framework is developed to enable the utilization of prosumers’ flexibility in the "control energy market" and "spot market" through flexibility aggregators. The change in prosumers’ pattern of power or energy consumption is within a specific range of services, where the loads (the demand side of the power grid) play a vital role by contributing to demand response (DR) programs. In this regard, first, the lowest layer of the proposed hierarchical energy management system has been synthesized for the two main applications. Then, the core energy management systems of the use cases at this layer have been extended to a unified and more generic hierarchical framework with the capability to be adapted to the two use cases to enable the prosumers for DR service provision in different time scales. However, this generic structure can be adapted to any prosumer and is not necessarily limited to the use cases explained in this thesis. In the use case of beverage production plants at the lower layer, a centralized MPC for economically optimal energy consumption and DR control of the plants has been suggested to assign the processing speed of the units inside the lines using a mixed-integer nonlinear program (MINLP). Also, the dynamical models of the processing units in the production lines have been considered for design purposes. In the use case of the bidirectional EV charging hub, the lower layer has been proposed in two different structures to optimally manage energy transactions between the charging hub (equipped with an energy buffer), the grid, and the EVs while considering DR and drivers’ preferences. Mixed integer optimization frameworks have been proposed for decision-making on charging or discharging the EVs. Finally, the generic hierarchical structure has been established and adapted to the use cases while exploiting mixed-integer optimization frameworks accordingly. The effectiveness of the proposed controllers for each use case has been investigated through various simulation scenarios. The solvers, such as IPOPT and Gurobi, have been exploited as the underlying tools for solving the MPC problems.