Efficient Customer Selection Process for Various DR Objectives

For timely and effective dispatch of demand response (DR) events, utilities require an efficient customer selection process that considers multiple factors, such as customer energy consumption patterns, customer compliance, and DR event time intervals. Moreover, customer targeting strategies may be different depending on the priority of each DR program. Thus, this paper investigates how to efficiently select customers for certain types of incentive-based DR programs (e.g., direct load control, curtailable load control) with various objectives based on hourly energy consumption data. To cope with computation issues arising from big data, we propose a filtering process that reduces the number of eligible customers based on proper criteria and use a novel approximate algorithm to solve the proposed customer selection problems in the format of a stochastic knapsack problem (SKP). This paper also suggests how to consider a customer compliance factor in the customer selection process.