An energy management strategy (EMS) is important for hybrid electric vehicles (HEVs) since it plays a decisive role on the performance of the vehicle. However, the variation of future driving conditions deeply influences the effectiveness of the EMS. Most existing EMS methods simply follow predefined rules that are not adaptive to different driving conditions online. Therefore, it is useful that the EMS can learn from the environment or driving cycle. In this paper, a deep reinforcement learning (DRL)-based EMS is designed such that it can learn to select actions directly from the states without any prediction or predefined rules. Furthermore, a DRL-based online learning architecture is presented. It is significant for applying the DRL algorithm in HEV energy management under different driving conditions. Simulation experiments have been conducted using MATLAB and Advanced Vehicle Simulator (ADVISOR) co-simulation. Experimental results validate the effectiveness of the DRL-based EMS compared with the rule-based EMS in terms of fuel economy. The online learning architecture is also proved to be effective. The proposed method ensures the optimality, as well as real-time applicability, in HEVs.

In order to mitigate the power density shortage of current energy storage systems (ESSs) in pure electric vehicles (PEVs or EVs), a hybrid ESS (HESS), which consists of a battery and a super-capacitor, is considered in this research. Duo to the use of the two ESSs, an energy management should be carried out for the HESS. An optimal energy management strategy is proposed based on the Pontryagin's Minimum Principle (PMP) in this research, which instantaneously distributes the required propulsion power to the two ESSs during the vehicle's propulsion and also instantaneously allocates the regenerative braking energy to the two ESSs during the vehicle's braking. The objective of the proposed energy management strategy is to minimize the electricity usage of the EV and meanwhile to maximize the battery lifetime. A simulation study is conducted for the proposed energy management strategy and also for a rule-based energy management strategy. The simulation results show that the proposed strategy saves electricity compared to the rule-based strategy and the single ESS case for the three typical driving cycles studied in this research. Meantime, the proposed strategy has the effect of prolonging the battery lifetime compared to the other two cases.

Regenerative braking is an effective approach for electric vehicles (EVs) to extend their driving range. A fuzzy-logic-based regenerative braking strategy (RBS) integrated with series regenerative braking is developed in this paper to advance the level of energy-savings. From the viewpoint of securing car stability in braking operations, the braking force distribution between the front and rear wheels so as to accord with the ideal distribution curve are considered to prevent vehicles from experiencing wheel lock and slip phenomena during braking. Then, a fuzzy RBS using the driver’s braking force command, vehicle speed, battery SOC, battery temperature are designed to determine the distribution between friction braking force and regenerative braking force to improve the energy recuperation efficiency. The experimental results on an “LF620” prototype EV validated the feasibility and effectiveness of regenerative braking and showed that the proposed fuzzy RBS was endowed with good control performance. The maximum driving range of LF620 EV was improved by 25.7% compared with non-RBS conditions.

Battery energy storage management for electric vehicles (EV) and hybrid EV is the most critical and enabling technology since the dawn of electric vehicle commercialization. A battery system is a complex electrochemical phenomenon whose performance degrades with age and the existence of varying material design. Moreover, it is very tedious and computationally very complex to monitor and control the internal state of a battery’s electrochemical systems. For Thevenin battery model we established a state-space model which had the advantage of simplicity and could be easily implemented and then applied the least square method to identify the battery model parameters. However, accurate state of charge (SoC) estimation of a battery, which depends not only on the battery model but also on highly accurate and efficient algorithms, is considered one of the most vital and critical issue for the energy management and power distribution control of EV. In this paper three different estimation methods, i.e., extended Kalman filter (EKF), particle filter (PF) and unscented Kalman Filter (UKF), are presented to estimate the SoC of LiFePO₄ batteries for an electric vehicle. Battery’s experimental data, current and voltage, are analyzed to identify the Thevenin equivalent model parameters. Using different open circuit voltages the SoC is estimated and compared with respect to the estimation accuracy and initialization error recovery. The experimental results showed that these online SoC estimation methods in combination with different open circuit voltage-state of charge (OCV-SoC) curves can effectively limit the error, thus guaranteeing the accuracy and robustness.

In order to realize the online learning of a hybrid electric vehicle (HEV) control strategy, a fuzzy Q-learning (FQL) method is proposed in this paper. FQL control strategies consists of two parts: The optimal action-value function Q*(x,u) estimator network (QEN) and the fuzzy parameters tuning (FPT). A back propagation (BP) neural network is applied to estimate Q*(x,u) as QEN. For the fuzzy controller, we choose a Sugeno-type fuzzy inference system (FIS) and the parameters of the FIS are tuned online based on Q*(x,u). The action exploration modifier (AEM) is introduced to guarantee all actions are tried. The main advantage of a FQL control strategy is that it does not rely on prior information related to future driving conditions and can self-tune the parameters of the fuzzy controller online. The FQL control strategy has been applied to a HEV and simulation tests have been done. Simulation results indicate that the parameters of the fuzzy controller are tuned online and that a FQL control strategy achieves good performance in fuel economy.

State of charge (SOC) estimation plays a critical role in the operation of an electric vehicle (EV) power battery. In this Letter, the authors propose machine learning (ML) algorithms as alternate to the existing filtering algorithms used for SOC estimation of an EV battery. The SOC estimation approach is evaluated by the simulation experiments in advanced vehicle simulator (ADVISOR). For the modelling of ML algorithms, the input parameters that affect the SOC estimation are battery current, battery module temperature, power out of the battery (available and requested), battery power loss and heat removed from the battery. Training and testing stages of the models are carried out using the data collected from ADVISOR. As the drive cycle conditions provided by ADVISOR are universal therefore present method is applicable to all kinds of batteries used in EVs including lithium ion, nickel metal hydride and lead acid batteries. Thus, the proposed models for SOC estimation provide an alternative approach in SOC estimation.