Power customer classification based on OCHNN-Kmeans and price setting of TOU

This paper starts from the power demand side, which proposes a time-sharing peak-valley pricing method for power users. Firstly, the OCHNN-Kmeans algorithm is adopted for the clustering of power load curves. This algorithm can overcome the disadvantages of K-means clustering instability. The use of OCHNN-Kmeans to cluster power consumer load data can meet the high-dimensional characteristics of power load data. On the other hand, the calculation speed of this algorithm is relatively fast and can be used in the practical application of power industry; According to the clustering results, typical user load data are extracted from each type of electricity consumers. The parameters of peak-to-peak, peak-flat, and valley-bottom load transfer rates were estimated using the constrained least-squares method. The estimated load value after the peak-to-valley time-of-use electricity price is changed, the estimated load value is fine-tuned in combination with the electricity price elasticity matrix, and the design optimization example is used to formulate optimal time-average peak-to-valley prices for different classes; finally, the domestic south is combined. An empirical analysis of the real data of a city provides that the development of specific time-of-the-hour peak-valley tariffs for different power users can reduce the electricity cost of demand-side power users, enable peak load-filling of load curves, and increase the efficiency of energy use. Meet the diverse needs of power users.