An online updating framework of an energy management system (EMS) for a multimode hybrid electric powertrain is proposed via cooperation between the asynchronous advantage actor–critic (A3C)-based deep reinforcement learning (DRL) agent and the Markov chain model (MCM). In the overall framework, the DRL agent periodically updates the energy management policy. The MCM expedites the policy update process by generating plenty of probable future drive cycles using recent historical driving data and supplying them to the training process. Assisted with the MCM, the proposed A3C-based energy management framework can yield near-optimal policy for any type of unknown drive cycle in the recent future. Two types of unknown drive cycles are chosen to demonstrate the efficacy of the proposed framework. Type I unknown drive cycle is also generated from the same recent historical driving data but was not included in the training dataset. Type II drive cycle is neither known to the framework nor generated from the same historical data. In type I unknown drive cycle, the trained A3C-based EMS achieves 99% of the fuel economy obtained by the global-optimal EMS and 0.12% deviation from charge sustainability. The trained A3C-based EMS consumes 6%–12% more fuel than the global-optimal EMS for type II unknown drive cycles.

Recent studies have addressed the development of optimal control strategies for hybrid electric vehicles (HEVs). Achieving global optimality for the fuel economy prediction while minimizing the computational efficiency still is a research and development challenge. This paper aims at presenting a novel technique for managing the energy flows in a power split HEV named slope-weighted energy-based rapid control analysis (SERCA). After presenting the HEV plant model and the optimal control problem, the currently most adopted energy management strategies are analyzed. The SERCA technique is then illustrated and its operating steps are detailed. The simulation results for the considered HEV energy management strategies in the standard urban driving cycle subsequently indicate that the SERCA can efficiently achieve near-optimal fuel economy while limiting the computational costs. This suggests the potential use of SERCA for rapid component sizing of HEV powertrains.

The multimode power-split architecture for hybrid electric vehicle (HEV) powertrains is generally known for the complexity of its operation. This paper first addresses the challenge of developing an automated on-line near-optimal control strategy for these systems. Particularly, a machine learning logic based on supervised learning is developed for on-line selection of the HEV operating mode, thus the particular set of clutches to be engaged. An efficiency-based approach is then adopted to determine the optimal power-split between powertrain components. Later, the developed strategy finds integration in an optimal design methodology for multimode power-split HEVs considering the effectiveness and ease of on-line controllability. The obtained results are compared with the ones by the traditional HEV design methodology that only considers off-line energy management. The illustrated design methodology with on-line control reveals efficient at identifying the multimode HEV design that demonstrates optimal predicted fuel economy values and ease of on-line controllability simultaneously. Results suggest that the HEV design optimization procedure may produce different outcomes and demonstrate more effective when the evaluation of the on-line controlled operation is integrated.

This work aims at presenting a design methodology capable of modeling, generating, and testing a large number of multimode power split hybrid electric vehicle transmission designs in a relatively short period of time. Design parameters include the planetary gear ratios, the final drive ratio, the configuration of hookups to link the hybrid powertrain components to the planetary gear sets and the locations of clutch connections between different nodes of the planetary gear sets. The system modeling approach is first presented, including formulations for each component (the vehicle and road load, the engine, the motor/generators and the battery). A rapid and automated modeling procedure is proposed for hybrid electric vehicle transmissions including multiple planetary gear sets and clutch connections. Two algorithms are subsequently presented that enable fast evaluation of fuel economy and acceleration performance of hybrid electric vehicle transmission designs, namely the enhanced Power-Weighted Efficiency Analysis for Rapid Sizing and the Rapid Efficiency-based Launching Performance Analysis algorithms. The developed design methodology is tested by first modeling and evaluating three hybrid electric vehicle designs from the state-of-art. Later, an investigation for optimal designs that can ameliorate the examined benchmarks is performed. Several millions of design options are rapidly generated and tested using the proposed procedure. The methodology is proved effective by quickly coming up with two sub-optimal designs. Fuel economy and acceleration performance are improved by 5.56% and 40.56%, respectively, compared to the corresponding best benchmarks.

Dynamic programming (DP) is currently the reference optimal energy management approach for hybrid electric vehicles (HEVs). However, several research concerns arise regarding the effective application of DP for optimal HEV control problems which involve a significant number of control variables, state variables and optimization constraints. This paper deals with an optimal control problem for a full parallel P2 HEV with constraints on battery state-of-charge (SOC), battery lifetime in terms of state-of-health (SOH), and smooth driving in terms of the frequencies of internal combustion engine (ICE) activations and gear shifts over time. The DP formulation for the considered HEV control problem is outlined, yet its practical application is demonstrated as unfeasible due to a lack of computational power and memory in current desktop computers. To overcome this drawback, a computationally efficient version of DP is proposed which is named Slope-weighted Rapid Dynamic Programming (SRDP). Computational advantage is achieved by SRDP in considering only the most efficient HEV powertrain operating points rather than the full set of control variable values at each time instant of the drive cycle. A benchmark study simulating various drive cycles demonstrates that the introduced SRDP can achieve compliance with imposed control constraints on battery SOC, battery SOH and smooth driving. At the same time, SRDP can achieve up to 78% computational time saving compared with a baseline DP approach considering the Worldwide Harmonized Light Vehicle Test Procedure (WLTP). On the other hand, the increase in the fuel consumption estimated by SRDP is limited within 3.3% compared with the baseline DP approach if the US06 Supplemental Federal Test Procedure is considered. SRDP could thus be exploited to efficiently explore the large design space associated to HEV powertrains.