Computation of privacy-preserving prices in smart grids

Demand management through pricing is a modern approach that can improve the efficiency of modern power networks. However, computing optimal prices requires access to data that individuals consider private. We present a novel approach for computing prices while providing privacy guarantees under the differential privacy framework. Differentially private prices are computed through a distributed utility maximization problem with each individual perturbing their own utility function. Privacy concerning temporal localization and monitoring of an individual's activity is enforced in the process. The proposed scheme provides formal privacy guarantees and its performance-privacy trade-off is evaluated quantitatively.