As an effective way to promote the usage of electric vehicles (EVs) and facilitate the consumption of distributed energy, the optimal energy dispatch of photovoltaic (PV) and battery energy storage systems (BESS) integrated fast charging stations with vehicle‐to‐grid is of considerable value to the efficient use of renewable energy. Based on an examination of the electrical structure and operation modes of PV and BESS integrated fast charging stations, considering the randomness of EVs’ arrival and departure, a rolling optimization strategy is adopted. Combining the energy architecture characteristics of the charging stations, the economic dispatch model is established by scheduling the orderly charging and discharging of EVs and the BESS to minimize the expense of daily operations. In view of the shortcomings of the first‐stage economic dispatch, two further optimization strategies are proposed to level out the power fluctuation of the grid connection point (POC) of the charging station, and the effect of the two‐stage optimization model and the multi‐objective optimization model are studied in comparison. Finally, an intraday real‐time optimization algorithm is designed, and the similarities and differences of the economic index and power fluctuation index are comprehensively analysed by using an actual charging station as a case study.