In this paper, a photovoltatic (PV) farm construction plan for charging plug-in electric vehicles (PEVs) in a workplace parking lot is proposed. The system involves a developed smart power charging controller, an under construction PV farm and a DC distribution bus connected to the AC utility grid. In order to produce enough energy for the daily consumption of a certain number of PEVs at the same time limit the construction cost of the PV farm, the scale of the PV farm should be optimized. A stochastic model of PEVs daily energy consumption is used together with the historical solar irradiance data in this area to find the best scale of the PV farm through genetic algorithm (GA) optimization. Since the output power from the PV farm varies in different environmental situations, energy forecasting and distribution is achieved by using the smart power flow controller. One day period simulation of the hybrid PEVs charging system is done to demonstrate that the proposed plan can effectively support enough power for a certain number of PEVs with the maximum utilization of renewable energy and minimum grid impact.