1 August 2018 in IEEE Systems Journal
This paper develops a decentralized competitive charging coordination algorithm for a large population of plug-in electric vehicles (PEVs) using the concept of a mean-field (MF) game. The aim of each PEV is to find its optimal charging strategy by minimizing an objective function consisting of charging cost, battery degradation cost, and benefit from charging, subject to the input and state constraints. The strategy of a PEV affects the objective functions of other PEVs through the electricity price, and therefore, we can model the interactions among PEVs as a game problem. No information exchange is considered among the PEVs. Each PEV only sends its own control action value to a population coordinator, and the coordinator just broadcasts a common signal to all the PEVs. This common signal is an estimate of the average control actions of PEVs and is called the MF term. Utilizing an adjustment mechanism for the MF term, a decentralized MF-optimal control algorithm is proposed, and it is shown that the algorithm converges to the $\varepsilon _N$ -Nash equilibrium point of the game, with $\varepsilon _N$ uniformly converging to zero as the population sizes of the PEVs go to infinity. Simulation results and comparison with other methods are performed to clarify the advantages of the proposed method.
18 October 2018 in IEEE Transactions on Intelligent Transportation Systems
Perimeter traffic flow control, based on the macroscopic fundamental diagram (MFD), has been introduced for traffic control and congestion management in large-scale networks. The perimeter controller is a set of traffic signals on the border between the regions manipulating the transfer flows with the aim to maximize the number of trips that reach their destinations. This paper tackles the optimal perimeter control of MFD systems for two-region urban networks which model heterogeneously congested cities. The modeling of the system results in nonlinear state dynamics, a non-quadratic cost function, and constraints on control actions and traffic states. We prove the existence of the optimal controller, analytically derive the optimal control policy, and introduce a numerical method to solve the optimal control policy. Based on the indirect optimal approach, HJB equation, and Pontryagin's maximum principal, we demonstrate that the optimal controller is in the form of Bang-Bang control. We apply the Chebyshev pseudospectral method to solve the two-point boundary value problem (TPBVP) for the proposed constrained optimal control problem. Consequently, the TPBVP is reduced to determination of the solution of a nonlinear system with algebraic equations. A numerical study is performed to measure the effectiveness of the proposed method.
23 June 2017 in IEEE Transactions on Automatic Control
In this paper, the theory of robust min-max control is extended to hierarchical and multiplayer dynamic games for linear quadratic discrete time systems. The proposed game model consists of one leader and many followers, while the performance of all players is affected by disturbance. The Stackelberg-Nash-saddle equilibrium point of the game is derived and a necessary and sufficient condition for the existence and uniqueness of such a solution is obtained. In the infinite time horizon, it is shown that the solution of the Riccati equation is upper bounded under a condition that is called individual controllability. By assuming such a condition and using a time-varying Lyapunov function the input-to-state stability of the hierarchical dynamic game is achieved, considering the optimal feedback strategies of the players and an arbitrary disturbance as the input.
22 October 2019 in IEEE Transactions on Sustainable Energy
30 October 2018 in IEEE Transactions on Vehicular Technology
This paper addresses the problem of energy consumption management of a large population of plug in hybrid electric vehicles (PHEVs). A PHEV can be derived either on its gasoline or the battery modes while traveling, and can be either charged from or discharged to the grid while parking at home, respectively. This feature is the main novel consideration in our system model, besides taking into account the constraints like traveling time and distance, and the bounds on the charging power and state of charge (SoC) of the battery. On the basis of the proposed model of PHEV, the optimal battery-gasoline energy management strategy is obtained in order to minimize the cost of energy and battery degradation for each user. Such an strategy can also compensate the peak load. Since the strategy of each PHEV affects other users' cost through the electricity price, we model such an interaction as a game problem and we prove that there exist a unique Nash equilibrium point of the game. Then, due to the fact that the effect of a single PHEV of a large population on the other users' costs is negligible, the theory of mean filed (MF) game is utilized to develop a decentralized control strategy for PHEVs. The lightweight information exchange of the proposed MF algorithm ensures a low computational cost and communication overhead. The proof of convergence for the proposed algorithm is given and it is shown that as the number of PHEVs goes to infinity the MF optimal strategy of the players converges to the unique Nash equilibrium point of the game. In simulation results, the effectiveness of the proposed method in terms of demand side management, peak shaving, convergence rate and robustness is studied.
6 May 2021 in IEEE Transactions on Automatic Control
We propose a two-layer, semidecentralized algorithm to compute a local solution to the Stackelberg equilibrium problem in aggregative games with coupling constraints. Specifically, we focus on a single-leader, multiple-follower problem, and after equivalently recasting the Stackelberg game as a mathematical program with complementarity constraints (MPCC), we iteratively convexify a regularized version of the MPCC as the inner problem, whose solution generates a sequence of feasible descent directions for the original MPCC. Thus, by pursuing a descent direction at every outer iteration, we establish convergence to a local Stackelberg equilibrium. Finally, the proposed algorithm is tested on a numerical case study, a hierarchical instance of the charging coordination problem of plug-in electric vehicles.