Optimal scheduling management of the parking lot and decentralized charging of electric vehicles based on Mean Field Game

Abstract

As an intermediary for the interaction between the grid and electric vehicles (EVs), the parking lot aggregator not only facilitates the exchange of electricity between EVs and the grid, but also brings benefits to all participants. This paper proposes a linear quadratic (LQ) Mean Field Game (MFG) theory with a major player to optimal scheduling management of the parking lot and formulate optimal decentralized charging control strategies for a large number of EVs, to achieve the minimization of EVs charging cost while maximizing the profit of the parking lot, but these two problems are a set of coupled control problems. In addition to modeling the interaction between the parking lot and the EVs as a finite-time dynamic game problem, the Nash Certainty Equivalence (NCE) of related optimization problems is also proposed, and the corresponding solution algorithm is designed. The control effects of the proposed dynamic game problem on the charging cycle, as well as the effects of parameters change and electricity price fluctuations on charging control are illustrated through numerical simulations.