Attack-Resilient Communication-Efficient Economic Dispatch Algorithm Design for Virtual Power Plants and Parameter Analysis

In this paper, we focus on the development of economic dispatch algorithms using communication-censoring strategy, applied to distributed generators (DGs) in virtual power plants (VPPs). Virtual power plants contain a large number of distributed generators, which offer many advantages compared to traditional power systems. However, this also makes them extremely susceptible to network-attacks. Additionally, under the condition of limited communication resources, there are issues such as communication congestion caused by excessive communication times. In order to improve the communication efficiency in virtual power plants economic dispatch, we first implement a distributed, communication-efficient dispatch algorithm called Co-DRA. However, Co-DRA cannot work properly under network attack. Therefore, taking into account the communication cost and security issues, we implement a communication-efficient attack-resilient economic dispatch algorithm called Co-BREDA. The feature of Co-BREDA is that if the local information an agent needs to transmit differs little from the last transmission, then this agent will not perform iteration in this round, and by means of this communication-censoring strategy, it reduces the number of communication times between agents. Theoretical analysis and numerical experiments on the IEEE-118 bus power system show that, with the reasonable selection of the threshold function, compared with the existing algorithm, Co-BREDA can significantly improve communication efficiency while maintaining comparable accuracy to the original algorithm, thus having strong practical significance.