Game theoretic approach for electrified auxiliary management in high voltage network of HEV/PHEV

Abstract

Auxiliary electrification becomes a potential solution to reduce the vehicle energy consumption. However, electrified auxiliaries operate mostly in individual way, non-cooperative in regardless of the vehicle state. In this paper, a new control strategy for electrified auxiliary system is proposed in order to improve the coordination among auxiliaries. This new control strategy is not only based on a game theoretic approach but also a model predictive control (MPC). In this approach, each electrified auxiliary is considered as a player participating in an energy consumption game, where players have incentive to cooperate and improve the global vehicle consumption. Simulation results on a plug-in hybrid electric vehicle show that this new control design provides a promising and simple approach to control the electrified auxiliary system.

Complete vehicle energy management using adaptive primal–dual operator splitting

“…CVEM requires highly flexible and scalable optimization methods (e.g., subsystems should be easily added or removed) and the optimization methods cited above do not have this property. Therefore, distributed (convex) optimization methods have been studied in East and Cannon (2019), Nguyen et al (2014), Romijn et al (2019), in which East and Cannon (2019) use the alternating direction method of multipliers, for solving drive cycles of up to 1000 time steps, and Nguyen et al (2014) use game theory in which limited data exchange is needed, but for which global optimality might not be achieved.…”

Real-Time Distributed Economic Model Predictive Control for Complete Vehicle Energy Management

“…Solving (26) subject to (27) and (28a) is hard due to the nonlinear vehicle dynamics, the appearance of the derivative dv dt and the constraints (28) and because the road slope is given as function of distance and not as function of time. Therefore, we propose to make a forward Euler approximation of (26) over the time horizon…”

“…The dynamics of the air pressure in the lumped system are assumed to satisfy a mass balance (see, e.g., [28]) given by:…”

“…Rule-based strategies require a new set of rules for every subsystem that is added to the CVEM problem; ECMS requires tuning of an equivalence factor for each additional state in the CVEM problem; and MPC requires reformulating the optimal control problem, the cost function and tuning of the penalizing parameters each time a subsystem is added. To enhance the flexibility of vehicle energy management, a distributed offline solution strategy is proposed in [26], and online solutions are proposed in [27][28][29]. In [27], an online game-theoretic approach to CVEM is presented, but prediction information is not taken into account.…”

“…In [27], an online game-theoretic approach to CVEM is presented, but prediction information is not taken into account. In [28], an online game-theoretic approach in combination with MPC is presented, which does utilize prediction information, but requires solving a nonlinear program. In [29], flexibility is obtained by using the Alternating Direction Method of Multipliers (ADMM), while ideas based on ECMS are used to provide the equivalent costs at a supervisory level.…”