Cooperative trading in smart-grid networks

Abstract

Continuous Double Auction (CDA) methods provide an efficient mechanism for price matching between buyers and sellers. Hence, these techniques have been adopted in smart-grid and micro-grid scenarios for Peer-to-Peer (P2P) energy trading. However, existing CDA methods either violate the natural ordering of P2P trading or introduce considerable overhead or make unrealistic assumptions about the underlying communication setup. In this paper, we introduce a CDA method involving trading agents that are coordinated with token allocation system for participation in the smart-grid market. Our proposed trading mechanism minimizes the natural order violations in the trading, and can recover from issues like token loss arising due to node or network connectivity issues. We provide algorithms for different operations of the CDA such as token request, token allocation, trading, etc. The proposed CDA mechanism has been evaluated with publicly available energy demand data and compared with another CDA method to show that it reduces natural ordering violations. Subsequently, we describe a variant of the CDA method where bids/asks from different participants are queued up for potential future price matches, thereby elaborating the scope for P2P trade. Both CDA variants can scale to large number of participants and can also work with heterogeneous bidding strategies adopted by clients.