Load curve clustering method based on time sequence variable density processing means

The invention discloses a load curve clustering method based on a time sequence variable density processing means. The load curve clustering method comprises the following steps: determining peak, valley and climbing periods of a load according to a load curve; carrying out density reduction processing on the load by adopting a self-adaptive segmented aggregation approximation method; determining data points corresponding to peak, valley and climbing periods after density reduction, and for the load sequence after density reduction, performing density increase processing on the peak, valley and climbing periods of the load by adopting a linear interpolation method; for the load sequence after density reduction and density increase, a similarity index is constructed by adopting a comprehensive criterion of Euclidean distance and related distance, and the similarity between loads is calculated; and for the load sequences after density reduction and density increase, based on the similarity between the loads, a