Dynamic Energy Trading and Load Scheduling Algorithm for the End-User in Smart Grid

Abstract

In the smart grid, the end-users have the opportunity to integrate renewable energy sources (RESs) and participate in two-way energy trading. At the same time, an increasing number of flexible loads (FLs) have been developed for use on the demand side. Thus, this paper considers joint energy trading and load scheduling at a end-user with integrated renewable generation. With unknown statistics on renewable generation, loads and electricity prices, we aim at optimizing both energy trading and load scheduling to maximize the long-term average profit of the end-user, subject to load delay constraints. We employ the Lyapunov optimization theory to solve this stochastic problem with a series of problem corrections and transformations that enable us to design a dynamic energy trading and load scheduling algorithm. Within the performance analysis of the algorithm, we further demonstrate that the algorithm not only provides a bounded performance guarantee to the optimal solution that has complete future information, but is also asymptotically equivalent to it as the battery capacity or the delay time of FLs tend to infinity. The simulation results show that the proposed algorithm is superior to other algorithms both in terms of performance and service delay. Moreover, we can achieve a trade-off between comfort and total profit by adjusting the values of the parameters, and analyze the effect of battery capacity on algorithm performance to provide a theoretical basis for the end-user to determine battery capacity.INDEX TERMS Smart grid, renewable energy sources, energy trading, flexible loads, load scheduling.

A Distributed Online Algorithm for Promoting Energy Sharing Between EV Charging Stations

“…Given the current system state s(t), the offline optimization of the charging stations system casts down to a stochastic optimization that minimizes the time-average total operational cost, as follows: 4), ( 5), ( 8) − (11), ( 13), ( 16), ( 17) (23) where the expectation symbol E[•] in the objective function is with regard to the random system state and corresponding control decisions.…”

“…For example, Lyapunov optimization was used to improve PV consumption of a cluster of nanogrids [22]. The energy trading between an end-user and grid was optimally scheduled to maximize the profit via Lyapunov optimization [23]. However, the above works were based on the centralized scheme with concerns about privacy and communication burden.…”