This paper introduces a dual genetic algorithm (DGA)-based approach for a fast screening process for distributed generations (DGs) connected to a distribution network. The existing deterministic approaches overlook some operation conditions that may give rise to incorrect results and lead to wrong decisions in practical applications. The DGA is an extension of a traditional single genetic algorithm (GA). Owing to complicated distribution system operating states, such as the actual topologies and features, and the load states, it is hard to take all uncertainties into account when the latter is used. Hence, an improved GA-DGA-is proposed to deal with these kinds of uncertainties. The outcomes are of value for a fast screening process for grid interconnection applications in DGs.