In order to more accurately determine the site of a converter station and configure its capacity , the location and capacity selection of the station in a DC distribution network is studied. Using mathematical statistical methods to analyze the measured power data of photovoltaic cells and wind turbines, four-year output time-series curves of photovoltaic cells and wind turbines are obtained, and the distributed power source in the planning area is determined with the minimum distributed power access cost and load moment as the objective function’s position. The traditional load moment model is improved, taking into account factors such as the size of the load in the planning area, the nature of the load, the simultaneous use of the load, and the geographical environment. At the same time, considering the seasonal fluctuation of distributed power sources, an improved load moment model is proposed. A genetic algorithm is used to optimize the model to determine the site and capacity of the converter station. This paper can be used as a reference for related engineering and technical research. This work is supported by National Natural Science Foundation of China (No. 51777142) and Science and Technology Project of State Grid Henan Electric Power Company (No. 5217L018000N).