Line loss analysis method based on self-organizing mapping neural network

The invention discloses a line loss analysis method based on a self-organizing mapping neural network. The method comprises the following steps: collecting the power utilization data of an original power grid line, and the power utilization data of the original power grid line comprises the power utilization data of a plurality of lines; preprocessing the original power utilization data of the power grid line; extracting power utilization characteristics of the line based on the preprocessed power utilization data of the original power grid line; extracting time sequence statistical characteristics based on the line power utilization characteristics; performing standardization processing on the time sequence statistical characteristics and the preprocessed power consumption data of the original power grid line, and performing integration to obtain to-be-analyzed total data; and constructing a line loss analysis model, inputting the to-be-analyzed total data into the line loss analysis model, and obtaining a l