Predictive Energy Management Methods for Smart Grids
Hooshmand, Ali 1982-
MetadataShow full item record
In this dissertation, we propose energy management methods for power systems in the context of smart grids. In this regard, we consider new management problems for various conﬁgurations of smart grids, microgrids, as well as the power system generation. For diﬀerent scenarios, we consider grid connection and distributed generations such as photovoltaic cells, wind turbine, and microgas turbines as energy sources. In addition, the eﬀects and advantages of storage devices in smart grids operation are investigated by including them as one of the system components. For microgrids operation, we consider a microgrid both in islanded mode and grid-tied mode of operation. In these modes, we develop and solve new optimization problems which aim to minimize the cost of energy within a microgrid to supply the load and maximize the lifetime of battery units simultaneously. Next, we extend the concept and consider a network of microgrids which are able to collaborate with each other. By proposing a cooperative optimization problem for microgrids network, we will show that the total cost of energy would be minimized. On the generation side, we investigate the economic dispatch problem for power systems which include renewable sources among energy providers. In this case, we will illustrate that conventional approaches for considering renewable energy sources in the dispatching problem will not be functional anymore. In addition, we will develop a new method which can be an appropriate alternative for conventional approach. Finally, we will investigate the advantages of storage devices in the aforementioned economic dispatch problem. Model predictive control (MPC) policies, in both deterministic and stochastic forms, are employed to solve the underlying optimization problems. Several solution methods such as stochastic dynamic programming, linear programming, etc., will be employed to solve the MPC optimization problems. Numerous testbeds and experimental data including IEEE 14-bus system and California ISO data will be utilized to demonstrate the eﬃciency and optimality of the proposed energy management methods.