Abstarct: The thesis focuses on providing an efficient frxamework for implementing the problem of determining the optimal investment strategies of generation companies in restructured power systems. The proposed frxamework is modeled as a two-level optimization problem. At the upper level of the proposed frxamework, each generation company competes with its rivals to determine the optimal strategies to expand its power generation capacity by installing new units in candidate locations with the aim of maximizing profits. At the lower level of the proposed frxamework, each generation company competes with its competitors to sell electricity in a pool-baxsed electricity market. More precisely, at this level, the clearing process of the electricity market is modeled. In this thesis, the supply function equilibrium model is used to implement and examine the competition between different generation companies. This model is a combination of Cornet (quantity-baxsed competition model) and Bertrand (price-baxsed competition model). In order to increase the flexibility of the proposed frxamework, the uncertainty of decisions made by rivals of a particular generation company to invest in order to expand their production capacity, the uncertainty associated with the bidding of the competitors of this company to sell electricity, the amount and prices of electrical loads that offer in the electricity market are also considered. In order to model and accurately investigate the effect of these uncertain parameters, a two-point estimate method has been used. In order to solve the proposed frxamework, in the first step, the two-level optimization problem has converted to a single-level optimization problem with stochastic equilibrium constraints. This single-level problem has nonlinear parts that by linearizing these exxpressions, the problem eventually converted to a single-level mixed-integer linear optimization problem that allows the use of mathematical solvers. To determine the optimal investment strategies of the generation company in order to develop its production capacity in deregulated environments, the proposed frxamework has been implemented on the IEEE 24-Bus RTS test network and under various study cases and scenarios.