As the development of distributed photovoltaic (PV), the power flow distribution and the voltage level of the distribution network change dramatically, which increases the difficulty and complexity of the distribution network regulation. In this paper, a complex affine distribution power flow considering the power correlation of distributed photovoltaic generations is proposed, which improves the accuracy and effectiveness of power flow calculation of the distribution network. Firstly, the k-means clustering algorithm is used to divide the distributed photovoltaic into several clusters, and the complex affine model of distributed generations (PVs) and energy storage system (ESS) is established. The forward-backward sweep power flow based on the complex affine arithmetic is used to solve the problem. Compared with other algorithms, the completeness and accuracy of the proposed algorithm are verified. Two different scenarios are formed by establishing the complex affine power model of the energy storage element before and after accessing ESS. By comparison, the effect of energy storage on the reduction of distributed PV fluctuations is verified.