Equivalent Consumption Management Strategy (ECMS), a representative energy management strategy for hybrid electric vehicles (HEVs) derived from Pontryagin's minimum principle, is known to produce a near-optimal solution if the costate or equivalent factor of electric use is appropriately determined according to the driving conditions. One problem when applying the control concept to real-world scenarios is that it is difficult to precisely evaluate the performance of the control parameter before driving is complete, so the costate cannot be determined properly. To address this issue, this study proposes a practical method for estimating an appropriate costate based on Deep Q-Networks (DQNs), which is a reinforcement learning algorithm that uses a Deep Neural Network to evaluate the performances and determine the best control parameter or costate. The control concept benefits vehicle energy management by selecting the control parameter most related to stochastic conditions or future driving information based on artificial intelligence (AI), while optimal control is deterministically conducted by ECMS if the control parameter is given. Simply, only the implicit part of the optimal controller is solved via artificial intelligence. In the simulation results, not only does the proposed control concept outperform an existing ECMS that uses an adaptive technique for determining the costate, but the concept is also very feasible, in that it does not need a model for evaluating the performances.

The fuel efficiencies of Hybrid Electric Vehicles (HEVs) depend heavily on the control concepts, especially on energy management strategies, because the system efficiency of the vehicle is determined by the exploitation of the electric energy. Control concepts based on optimal control theories like Dynamic Programming (DP) or Pontryagin's Mminimum Principle (PMP) have been widely studied over the last two decades, and it has been proven that these control concepts provide near-optimal solutions if future driving conditions are given prior to actual driving. However, it is difficult to apply to real vehicles unless the future driving conditions are known. In this paper, a well-known control strategy based on PMP has been developed, which needs only current driving conditions. The significance of the study is that the control concept introduced herein was evaluated in the IEEE VTS Challenge 2018 competition. It produced the best results in terms of the total energy consumption, where 52 controllers had been competing to minimize the energy consumption in unknown driving cycles. Although the control system was designed from traditional PMP-based control, it becomes a practicable and effective solution by using an adaptive concept for balancing State Of Charge (SOC) of a battery. This study is an extended version of the study presented during IEEE VPPC 2018 and provides the development process of the control concept used in the controller during the IEEE VTS Challenge 2018.

Hybrid Electric Vehicles (HEVs) have become a mature technology for saving fuel and solving environmental problems. Fuel consumption in HEVs depends heavily on the control concept that distributes power among the power sources and manages the State of Charge (SOC) of the battery. Optimal control strategies based on the Equivalent Consumption Minimization Strategy have been thoroughly studied previously, and prior results show that the control strategies achieved high fuel efficiencies. However, predictions of driving patterns are required in order to implement the optimal control concept. Furthermore, obtaining driving patterns is very difficult because various driving situations should be considered when evaluating the equivalent fuel consumption of an HEV. Thus, the optimal control is difficult to implement in real-world applications. In this paper, an adaptive optimal control strategy based on Pontryagin''s Minimum Principle is introduced, which can be applied to real vehicles and does not require forecasting of driving patterns. Instead, it uses an adaptive concept for balancing the SOC. The performance of the control concept is evaluated in simulations for GM Volt 1st Gen, and the results show approximately 9.16% improvement in fuel consumption compared to other control concepts.