Asynchronous distributed cooperative energy management through gossip-based incremental cost consensus algorithm

The energy management problem in smart grid is a complex optimization problem of a Cyber-Physical System. Distributed cooperative energy management algorithms have great potential to solve this class of problems. In addition to the synchronous distributed algorithms, asynchronous distributed algorithms are more flexible, robust to packet loss and do not require global clock synchronization. In this paper, we have extended the synchronous Incremental Cost Consensus (ICC) algorithm to a gossip-based asynchronous version. The new algorithm is able to converge to the optimal solution in a distributed fashion with pairwise information exchange between neighbors without the need for any global synchronizing clock. The characteristics of the asynchronous ICC algorithm can be controlled by tuning the weighting of the updating matrix. Several case studies with different system configurations have been used to discuss the characteristics of the proposed algorithm.