Abstract
Integrated use of electricity and heat is an effective way to improve energy efficiency, precipitating the advent of multi-energy systems. In a network of multi-energy systems connected via power electronics devices, the power flow through soft-open points can be directly controlled, which allows peer-to-peer (P2P) energy trading. This paper proposes an online distributed decision scheme for this scenario. We begin with the deterministic centralized model of P2P energy trading problem, aiming to minimize long-term energy cost. Then the problem is decomposed temporally and spatially. To handle uncertainty, Lyapunov optimization is employed to decompose inter-temporal constraints in storage operation model, thus the problem in each period can be solved independently, yielding an online prediction-free policy. It is proven that, by properly selecting parameters, online policy generates a strategy that satisfies inter-temporal constraints and enjoys guaranteed performance. To protect privacy, an alternating direction method of multipliers (ADMM) based algorithm is employed to decompose the one-time-slot problem spatially, so that the subproblem of each peer admits a solver-free solution. The P2P trading price is intrinsically generated during the consensus-reaching process. Case studies demonstrate the prediction-free, performance-competitive, solver-free advantages of the proposed method. Both total and individual energy costs are reduced, and close to offline optimum.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call