Finite-Time Performance Mask Function-Based Distributed Privacy-Preserving Consensus: Case Study on Optimal Dispatch of Energy System

Abstract

Privacy-preserving consensus can address the information being leaked during distributed computing, encouraging its application in various scenarios. This paper investigates the finite-time privacy-preserving distributed optimal dispatch for energy systems (ESs). Firstly, a dynamic output mask function is designed to ensure that each node's internal state cannot be identified while accomplishing a distributed task. Second, two finite-time privacy-preserving consensus algorithms are presented, including leader–follower and average consensus algorithms. Under the proposed dynamic mask function, the proposed algorithms are local, allowing each node to protect its privacy by adopting the proposed dynamic output mask. The superiority of the proposed algorithm lies in its ability to achieve precise convergence while ensuring privacy protection. Third, the accurate value of the target state can be obtained after finite steps when processing and transmitting information. In addition, several conditions are presented for ensuring the convergence of the algorithms, which is not limited by special topologies such as undirected graphs and balanced graphs. Finally, an application that achieves the distributed optimal dispatch for the CCHP-based (Combined Cooling, Heating, and Power) ESs, and two examples illustrate that the algorithms can be effective access to economic optimization and excellent privacy performance.