The lack of available reserve for compensating imbalance between supply and demand is one of the important challenges facing the power systems with high penetration of renewables. However, the communication infrastructure in these systems is expanding, so as to provide a bidirectional channel for exchanging information and data between system operator, aggregators, and customers. Considering this capability beside the increasing number of electric vehicles (EVs) and controllable smart appliances in future smart grids, the idea of controlling the charge and discharge power of EVs together with intelligent loads to participate in frequency control has been proposed. In this thesis, to incorporate EVS into frequency control, a grouping method is proposed based on dividing the EVs into several groups by aggregators. Grouping is done using EVs information such as the current state of charge (SoC), the required SoC for the next trip, departure time, and the owner decision. In addition to using the results of calculations done by the aggregator in providing ancillary services to power system such as frequency control, they can be used to produce a dynamic model of EV used for frequency response studies. The scheme proposed in this thesis includes the determination of the primary reserve available from EVs and optimal tuning of load frequency control. To this end, several evolutionary algorithms are used to optimize load frequency controllers and their performance is compared. To evaluate the performance of the proposed frequency control scheme, and to study different conditions, four test systems including Taiwan power system, a three area power system, a micro-grid with different generating units and Spanish power system are used Keywords: Smart Grids, Electric Vehicles, Frequency Control, Primary and Secondary Reserve, Aggregate Model of Electric Vehicle, MicroGrids, Vehicle-to-Grid System.