Interest in using offshore wind power is rapidly increasing worldwide. A typical offshore wind farm may have hundreds of generators installed within an area ranging from several to tens of square kilometers. Therefore, many feasible schemes have been developed for optimizing wind turbine locations and internal line connections in a wind farm. The planner must search for the optimal scheme in terms of wind power output and installation and operation costs. This study proposes a novel procedure for optimizing wind turbine location and connection line topology by using artificial intelligent techniques: the Genetic Algorithm (GA) is used to optimize the layouts for the offshore wind farm, and Ant Colony System (ACS) algorithm is used to optimize line connection topology. The wake effect, real cable parameters and wind speed series are also considered. The proposed concepts and methods can improve the economy and efficiency of offshore wind farms worldwide.