Privacy Protection Decentralized Economic Dispatch Over Directed Networks With Accurate Convergence

Decentralized algorithms to solve the economic dispatch problem (EDP) in smart grids have been a significant focus within engineering research due to their advantages in scalability, robustness, and flexibility. Its purpose is to optimize the generation power of each generator to jointly achieve the minimal total generation cost on the premise of satisfying the total demand and generation capacity. Recently, the emergence of data security and the requirement for complex computing have led to a resurgence of activity in this area. To address EDP while considering the issues of private security and computation efficiency, we propose a novel privacy-protected decentralized random sleep algorithm over an unbalanced directed network. On the one hand, the proposed algorithm can effectively protect sensitive information by adding conditional noises in the state exchange. On the other hand, it can also promote computation efficiency over an unbalanced directed network by incorporating the random sleep strategy into the decentralized inexact gradient method with the gradient rescaling technique. It is proved that the proposed algorithm is able to achieve the optimal solution of the EDP. Furthermore, we also provide theoretical proof to guarantee the convergence and privacy properties of the proposed algorithm. Finally, two simulation examples of EDP in smart grids are provided to demonstrate the effectiveness of the proposed algorithm.