Nowadays, there is a conflict between the rapidly increasing demand for electricity and the requirement for reducing dependence on fossil fuel to decrease the greenhouse gas emissions. Proper utilization of renewable energy such as wind energy is proposed as an efficient solution to address this problem. However, due to the high inter-temporal variation and limited predictability, it is difficult to make full use of renewable energy as supplement to the conventional thermal power plants in smart grid. In this paper, we provide an energy efficient solution to solve this problem: the integration of Pug-In-Hybrid Electric Vehicle (PHEV). Through proper charging and discharging processes, the PHEV fleet can act as energy storage when there is excess renewable energy, while in case of energy shortage, the energy can be properly returned to the grid. Optimization models based on stochastic programming are developed for both the cases without PHEV fleet and with PHEV fleet, while various uncertainties such as power price, renewable power and user demand are taken into account. By solving the two stochastic programming problems, the optimal power management solutions are obtained, and the numerical results show that the integration of PHEV can effectively reduce the energy to be generated by conventional thermal power plants, and as a result, the overall energy generation cost can be significantly reduced.