Customers Baseline Load Estimation Based on Cluster Analysis of Control Group

Demand response (DR) improves the adjustment ability of power system from the user side, which plays an important role in improving the flexibility of power system and promoting the consumption of new energy. The estimation of customer baseline load (CBL) involves the benefits of DR participants and providers, so the accuracy of CBL estimation directly affects the implementation of DR projects. The control group method has a high accuracy. However, how to make full use of the non-DR load information of power users and improve the estimation accuracy of CBL is a problem that needs further research. This paper proposes a method for CBL estimation of power users based on k-means clustering control group. Firstly, the historical data of power users in the control group and the data on DR day are analyzed by k-means clustering to generate k load vectors of cluster center, and the historical data of users in the DR group in non-DR periods and the data on DR day are matched with the cluster center, and the weight vector was determined by the optimization method to estimate the CBL of DR users in DR periods. Taking the measured load data of residential users of a power company as an example, it is proved that the method can improve the estimation accuracy of CBL.