An Iterative Price-Based Approach for Optimal Demand-Response

Demand Response (DR) programs encourage consumers to adjust their power consumption in response to DR events such as changes in electricity prices or sudden peaks in demand. While significant savings can be thus achieved, the real success of DR programs depends on the incentive compatible participation of consumers and their timely response to DR signals. In this work, we consider households that operate different types of appliances consuming electricity loads characterized by preferences on the time and quantity of the consumption. We develop a distributed optimization algorithm of practical interest, which offers both to the providers and the consumers of electrical power the opportunity to jointly interact and converge to the optimal scheduling of their appliances: this optimizes the trade-off between user-savings with the inconvenience caused by shifting consumption in time and shading load, while leading the cost of the utility company to the socially optimal level. Our approach is based on related work, and solves two key practical issues: i) we convert the consumer reaction problem to the solution of an LP using standard LP solvers, and ii) we propose a heuristic for load consolidation, i.e., regroup appliance operation on adjacent timeslots, using a randomized algorithm. We also present a substantial amount of simulation data that suggests that our heuristics are sound and reach quickly a state relatively close to the actual system optimum. Keywords-Demand response; dynamic pricing; energy consumption scheduling; incentives; smart grid