Customer recognition-based demand response implementation by an electricity retailer

This paper introduces an innovative methodology based on clustering techniques to provide the retailer with a strategy to select the most suitable customers for implementing demand response programs (DRPs). The main aim is to minimize the cost of DRPs for supplying the demand during power shortage periods. For this purpose, customers with similar load profile are clustered in same cluster by using clustering techniques. Then, clusters following similar load pattern as the load curve of the retailer, especially during peak hours, are determined. In addition, a new concept is proposed to enable the customers to submit their demand reduction function based on the award offered by the retailer. A nonlinear optimisation approach is developed to minimize the cost function of the DRP and is solved using GAMS software. The proposed methodology is implemented on 114 customers of an electricity retailer in Tehran.