The invention provides a daily load characteristic curve extracting method based on an SOM neutral network clustering algorithm in order to overcome the defects of an existing daily load characteristic curve extracting method. The method includes the steps that firstly, daily load data of a user who can be monitored in a power consumer electricity utilization information collection system are acquired, so that daily load characteristic curve extraction and analysis are carried out; afterwards, normalization processing is conducted on the data, an input vector is stored, and a weight vector is initialized; at last, a daily load characteristic curve of the category user is obtained through weight correction, wherein a connection weight vector between a superior nerve cell and an input nerve cell is used as a clustering center in the daily load characteristic curve. The analysis shows that a good effect can be achieved for extracting the daily load characteristic curve by using the SOM neutral network clustering algorithm, the extracted daily load characteristic curve can well reflect the electricity utilization behavior and characteristic of the category user, and useful reference is provided for power enterprises to understand the electricity utilization habits of users, subdivide an electricity market, adjust a pricing strategy and implement demand side management.