Robust and Comfortable Day-Ahead Demand Side Management of Residential, Industrial and Commercial Consumers for Demand Normalization

Abstract

This paper proposes the robust and comfortable day-ahead demand side management (DSM) methodology with an aim to normalize the demand curve pertaining to residential, industrial and commercial consumers. The ideal normalized demand curve is one in which demand follows the inverse relation to the respective price. In this paper, the demand curve normalization problem is formulated as an optimization problem which minimizes the square of the error between the demand curve obtained after DSM and ideal demand curve. The constraints of the optimization problem include the modeling of demand shifting from high to low price periods, local generation and working habits of consumers. Due to the probabilistic nature of the forecasted day-ahead hourly prices, these can exhibit variability from the mean values; hence a robust optimization technique has also been incorporated such that the proposed DSM methodology remains robust under worst case of price variability. The simulation results are obtained (with varying degrees of discomfort and robustness) for an aggregator, which involves shiftable devices belonging to residential, commercial and industrial consumers. The intuitive and interesting benefits of the proposed methodology are presented and compared with existing approaches pertaining to the non-robust, un-comfortable, comfortable, demand flattening, and demand normalization based DSM.