Distributed Privacy-Preserving Algorithm for Economic Dispatch and Demand Response of Smart Grid With Homomorphic Encryption
Bing Liu, J.H. Liu T.B. Wu, Li Chai
Abstract
Recently, distributed privacy-preserving algorithms have drawn much attention in solving the economic dispatch and demand response (EDDR) problem of smart grids. However, existing privacy-preserving methods suffer from either limited protection performance (e.g., noise injection methods) or heavy computational complexity (e.g., cryptography-based methods). In this paper, we propose a distributed algorithm for the EDDR problem with satisfactory privacy preservation performance as well as modest computation complexity. In particular, the proposed algorithm integrates randomness into the weight matrix and seamlessly incorporates homomorphic encryption techniques to protect privacy from both honest-but-curious nodes and external eavesdroppers. Moreover, we address computational and communication overhead concerns by utilizing a single key pair for all nodes, encrypting only a portion of information on communication links, and minimizing information exchange between neighboring nodes to only once per iteration. Besides, we analyze and prove the convergence and privacy preservation of the proposed algorithm. Finally, we demonstrate the effectiveness by some examples, showing that the proposed algorithm effectively addresses the EDDR problem, while also providing better privacy preservation capabilities and reduced runtime compared to existing algorithms.