Residential demand response program: predictive analytics, virtual storage model and its optimization

Abstract

Demand response programs are becoming an integral part of the power system, helping create a closer alignment between the electrical service providers and customers. The research described in this paper uses the residential demand response (DR) program during a peak demand event to determine the demand reduction capacity as a virtual storage (VS). As in the marketing business, identifying target customers is vital in the DR program, thus making it more efficient and productive. Additionally, peak load events are very critical in the power system; therefore, it is essential to model an effective demand response program. This paper uses predictive analytics to estimate the level of residential participation in a DR program, and thus the load reduction capacity available, during peak load events. Also derives the mathematical modeling of the demand reduction capacity of a demand response program as a virtual storage system and optimizes it using genetic algorithm technique.