Power-split Plug-in Hybrid Electric Vehicle Research Articles

Bi-level energy management strategy for power-split plug-in hybrid electric vehicles: A reinforcement learning approach for prediction and control

The implementation of an energy management strategy plays a key role in improving the fuel economy of plug-in hybrid electric vehicles (PHEVs). In this article, a bi-level energy management strategy with a novel speed prediction method leveraged by reinforcement learning is proposed to construct the optimization scheme for the inner energy allocation of PHEVs. First, the powertrain transmission model of the PHEV in a power-split type is analyzed in detail to obtain the energy routing and its crucial characteristics. Second, a Q-learning (QL) algorithm is applied to establish the speed predictor. Third, the double QL algorithm is introduced to train an effective controller offline that realizes the optimal power distribution. Finally, given a reference battery's state of charge (SOC), a model predictive control framework solved by the reinforcement learning agent with a novel speed predictor is proposed to build the bi-level energy management strategy. The simulation results show that the proposed method performs with a satisfying fuel economy in different driving scenarios while tracking the corresponding SOC references. Moreover, the calculation performance also implies the potential online capability of the proposed method.

Comprehensive CO2 emissions analysis and optimization of plug-in hybrid electric vehicles

Fuel and grid have different life-cycle carbon emission intensities, which makes plug-in hybrid electric vehicles (PHEVs) have great low-carbon potential. To reduce carbon emissions during PHEV driving, the energy allocation of PHEV based on carbon intensity is studied. The energy life cycle carbon emissions factors are collected and energy management strategies (EMS) are developed. A power-split PHEV model is established to analyze carbon dioxide (CO2) emissions. Carbon emissions factors calculate the carbon emissions of each power flow. Simulation research is carried out using Tianjin and Yunnan’s energy structure as an example. The equivalent CO2 emission minimum strategy (ECEMS) is established by optimizing the hybrid mode in the Rule-based strategy. Based on ECEMS, the improved particle swarm optimization (PSO) algorithm is used to optimize the rules of hybrid mode switching. As a comparison, the same method but a different goal is used to optimize energy consumption. The results show that the lowest energy consumption does not mean the lowest carbon emission. CO2 emissions in different scenarios are reduced by 7.1%–31.7%, but the energy consumption is increased.

Electrothermal Dynamics-Conscious Many-Objective Modular Design for Power-Split Plug-in Hybrid Electric Vehicles

This article proposes an improved modular design methodology of a power-split plug-in hybrid electric vehicle (PHEV) that introduces an advanced electrothermal coupled model and a temperature-related subobjective to simultaneously reveal battery thermal and electrical dynamics in the modular design. Considering to provide customers with more optimal configuration solutions, a Pareto-augmented collaborative optimization (PACO) scheme is designed that integrates three benchmarking many-objective evolutionary algorithms (MOEAs) to expand the distribution of an approximated Pareto frontier composed of the best solution set. Two realistic worldwide harmonized light vehicles test cycles are separately reproduced by two trained drivers on a chassis dynamometer to test the robustness of the optimized vehicle system. The simulation results demonstrate that the MOEA based on decomposition (MOEA/D) in the PACO is the main contributor for PHEV modular design because it lessens the generational distance by at least 2.7% and enlarges the hypervolume by at least 17.6%, compared to the elitist nondominated sorting genetic algorithm and improved strength Pareto evolutionary algorithm. In the modular adaptation for different user types, the PHEV system optimized by the PACO can regulate cell temperatures ( <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$\mathbf{27}{{\bf .5}} - \mathbf{38}{{\bf .}}{\mathbf{3}^ \circ }\mathrm{C}$</tex-math></inline-formula> ) of all user types within a safe and efficient working zone ( <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$\mathbf{0} - \mathbf{5}{\mathbf{5}^ \circ }\mathrm{C}$</tex-math></inline-formula> ).

Co-Optimization of On-Ramp Merging and Plug-In Hybrid Electric Vehicle Power Split Using Deep Reinforcement Learning

Current research on Deep ReinforcementLearning (DRL) for automated on-ramp merging neglects vehicle powertrain and dynamics. This work considers automated on-ramp merging for a power-split Plug-In Hybrid Electric Vehicle (PHEV), the 2015 Toyota Prius Plug-In, using DRL. The on-ramp merging control and the PHEV energy management are co-optimized such that the DRL policy directly outputs the power split between the engine and the electric motor. The testing results show that DRL can be successfully used for co-optimization, leading to collision-free on-ramp merging. When compared with sequential approaches wherein the upper-level on-ramp merging control and the lower-level PHEV energy management are performed independently and in sequence, we found that co-optimization results in economic but jerky on-ramp merging while sequential approaches may result in collisions due to neglecting powertrain power limit constraints in designing the upper-level on-ramp merging controller.

Energy management strategy for power-split plug-in hybrid electric vehicle based on MPC and double Q-learning

Energy management strategy (EMS) for plug-in hybrid electric vehicle (PHEV) is the key to improve the energy utilization efficiency and vehicle fuel economy. In this paper, a model predictive control (MPC) based on EMS coupled with double Q-learning (DQL) is presented to allocate the power between multiple power sources for PHEV. Firstly, the powertrain framework of the PHEV and its mathematical models were analyzed in detail. Then, based on the required power and speed, an effective convergent offline learning controller was established based on DQL algorithm. Subsequently, the multi-feature input Elman neural network was implemented to predict vehicle speed in MPC, and the trained DQL controller was applied to solve the rolling optimization process in MPC to find the optimal battery output in the prediction horizon. Finally, the proposed strategy was verified in Autonomie software, and the simulation results show that the proposed strategy can achieve a superior fuel economy close to that of the offline stochastic dynamic planning strategy, meanwhile with a perfect adaptability for different state of charge (SOC) reference trajectories.

Dynamic Torque Coordinated Control Considering Engine Starting Conditions for a Power-Split Plug-In Hybrid Electric Vehicle

For a power-split plug-in hybrid electric vehicle (PS-PHEV), the control system cannot effectively reduce jerks under some engine-starting conditions due to the pulsating torque of the engine, elastic characteristics of the transmission, battery power limitations, and the limitation of the maximum torque of the motor. To solve this problem, a dynamic torque-coordinated control method considering engine starting conditions (DTCC-ESC) is proposed in this paper. The proposed DTCC-ESC mainly includes three parts: the engine segment active control before and after engine ignition to minimize the impact of the engine ripple torque on the shocks of the powertrain system; the feedforward control of the engine starting conditions and the feedback control of the engine’s optimal target speed; and the active damping feedback compensation control for system resonance to reduce system vibrations. The results show that the proposed DTCC-ESC can effectively improve driving comfort under different engine starting conditions.

Coordinated control strategy for mode transition of DM-PHEV based on MLD

In this paper, a hybrid variable cost function model predictive control strategy (HC-MPC) is developed for the mode transition process of a novel motor compound power-split plug-in hybrid electric vehicle (DM-PHEV). Based on the layout of DM-PHEV, the mode transition process from electric driving mode to hybrid driving mode is divided into 5 stages. Considering the hybrid characteristics of mode transition process, the dynamic models under different stages are transmitted into an equivalent mixed-logical dynamical hybrid model. MPC based on hybrid model (H-MPC) is applied as the coordinated controller. Moreover, considering that fixed cost functions for the finite time horizon are not reasonable if stage transition happened during the prediction horizon, so variable cost functions changing with predictable stages is useful for improving the control effect of MPC. H-MPC considering variable cost functions (HC-MPC) is proposed as the coordinate controller in this paper. Simulation and hardware-in-the-loop test show that HC-MPC can efficiently improve the PHEV drivability.

Extended Horizon ECMS Control of PHEVs With 2D Electricity Price Adaptation Policy

This article proposes new improvements to the ECMS-based control of plug-in hybrid electric vehicles (PHEVs) using macro-level trip forecast information. First, the dynamic programming (DP) optimization is applied to a power-split PHEV model to obtain the optimal powertrain trajectories for benchmarking the energy economy of the proposed controllers. An instantaneous energy cost minimization process is then formulated to compute the optimal mix of fuel and electricity, subject to the state and input constraints. To improve the speed and accuracy of the energy management system, the Nelder-Mead simplex search algorithm is used for the instantaneous optimizations. To account for the battery capacity limitation, an adaptive electricity pricing policy is introduced to control battery discharge rate based on the remaining trip distance, average speed, and elevation forecasts. The parameters of the proposed policy are optimized for a group of 5 urban and highway drive cycles to obtain a robust controller which can serve a wide spectrum of driving patterns. To further improve the energy economy, an extended-horizon optimization algorithm is proposed to account for control input transitions between consecutive time steps. Finally, a revised speed forecast formula is implemented during the low-SOC intervals to lift battery SOC ahead of high-demand periods. The combination of the proposed schemes enables reaching over 99% average optimality compared to a finely discretized DP optimization for the adopted drive cycles, and above 98% optimality for a test PHEV drive cycle.

Stochastic model predictive control for energy management of power-split plug-in hybrid electric vehicles based on reinforcement learning

In this paper, a stochastic model predictive control (MPC) method based on reinforcement learning is proposed for energy management of plug-in hybrid electric vehicles (PHEVs). Firstly, the power transfer of each component in a power-split PHEV is described in detail. Then an effective and convergent reinforcement learning controller is trained by the Q-learning algorithm according to the driving power distribution under multiple driving cycles. By constructing a multi-step Markov velocity prediction model, the reinforcement learning controller is embedded into the stochastic MPC controller to determine the optimal battery power in predicted time domain. Numerical simulation results verify that the proposed method achieves superior fuel economy that is close to that by stochastic dynamic programming method. In addition, the effective state of charge tracking in terms of different reference trajectories highlight that the proposed method is effective for online application requiring a fast calculation speed.

Energy Management for a Power-Split Plug-In Hybrid Electric Vehicle Based on Reinforcement Learning

This paper proposes an energy management strategy for a power-split plug-in hybrid electric vehicle (PHEV) based on reinforcement learning (RL). Firstly, a control-oriented power-split PHEV model is built, and then the RL method is employed based on the Markov Decision Process (MDP) to find the optimal solution according to the built model. During the strategy search, several different standard driving schedules are chosen, and the transfer probability of the power demand is derived based on the Markov chain. Accordingly, the optimal control strategy is found by the Q-learning (QL) algorithm, which can decide suitable energy allocation between the gasoline engine and the battery pack. Simulation results indicate that the RL-based control strategy could not only lessen fuel consumption under different driving cycles, but also limit the maximum discharge power of battery, compared with the charging depletion/charging sustaining (CD/CS) method and the equivalent consumption minimization strategy (ECMS).

Hierarchical Energy Management for Power-Split Plug-In HEVs Using Distance-Based Optimized Speed and SOC Profiles

This paper proposes a distance-based two-stage energy management strategy for power-split plug-in hybrid electric vehicles (PHEVs). One stage is for long-term optimization and the other is for short-term adaptation to actual traffic conditions. Energy consumption in PHEVs depends on the characteristics of the drivetrain as well as the operating conditions such as power demands and their split. Thus, prior to departure, the operating conditions for a whole trip are optimized for the drivetrain characteristics and trip information, which generates optimal speed and state-of-charge profiles. While driving, the operating conditions are adapted to current traffic conditions for a short horizon on the basis of long-term optimization results. In consideration of the changeability of traffic conditions, the proposed energy management strategy is performed in a distance domain, which localizes the effects of changes in traffic conditions on the long-term optimization results. Therefore, this distance-based two-stage strategy improves the balance between the optimality and the real-time computing time, which is suitable for online management. A model for the propulsion system in a PHEV and the energy management strategy were formulated in a distance domain. An estimation of distribution algorithm was used for long-term optimization and local adaptation.

Real-Time Fuel Economy Optimization With Nonlinear MPC for PHEVs

This brief addresses the energy management problem with the framework of receding horizon optimization. For power-split plug-in hybrid electric vehicles (HEVs), the real-time power-split decision is formulated as a nonlinear receding horizon optimization problem. Then, an online iterative algorithm to solve the optimization problem is proposed based on the continuation/generalized minimum residual algorithm. It should be noted that the proposed energy management strategy aims for optimality of the targeted horizon, but the solution is not optimal for the full driving route, unlike many solutions presented using the dynamic programming approaches. At each decision step, only the initial value of the optimal solution is implemented according to the receding horizon optimization approach. Finally, to demonstrate a comparison of the proposed scheme with other schemes, numerical validations conducted on a full-scale GT-SUITE HEV simulator are presented.

Component sizing of a power-split plug-in hybrid electric vehicle for optimal fuel economy

Plug-in hybrid electric vehicles (PHEVs) were introduced in response to rising environmental challenges facing the automotive sector. Of the three primary PHEV architectures, power-split architectures tend to provide greater efficiencies than the other ones; however, they also demonstrate more complicated dynamics. In this study, the problem of optimising the component sizes of a power-split PHEV was addressed in an effort to exploit the flexibility of this powertrain system and further improve the vehicle's fuel economy, using a Toyota plug-in Prius as the baseline vehicle. Autonomie software was used to develop the vehicle model. The engine's maximum power and the electric motor's maximum power were considered as the design variables. The genetic algorithm approach was employed to solve the optimisation problem. Comparing to the baseline vehicle, a significant reduction in fuel consumption was achieved thorough the sizing process for various drive cycles of FTP, HWFET and EPA. The model was validated against a MapleSim multi-domain model.

Dynamic Traffic Feedback Data Enabled Energy Management in Plug-in Hybrid Electric Vehicles

Recent advances in traffic monitoring systems have made real-time traffic velocity data ubiquitously accessible for drivers. This paper develops a traffic data-enabled predictive energy management framework for a power-split plug-in hybrid electric vehicle (PHEV). Compared with conventional model predictive control (MPC), an additional supervisory state of charge (SoC) planning level is constructed based on real-time traffic data. A power balance-based PHEV model is developed for this upper level to rapidly generate battery SoC trajectories that are utilized as final-state constraints in the MPC level. This PHEV energy management framework is evaluated under three different scenarios: 1) without traffic flow information; 2) with static traffic flow information; and 3) with dynamic traffic flow information. Numerical results using real-world traffic data illustrate that the proposed strategy successfully incorporates dynamic traffic flow data into the PHEV energy management algorithm to achieve enhanced fuel economy.

Multi-objective component sizing of a power-split plug-in hybrid electric vehicle powertrain using Pareto-based natural optimization machines

The urgent need to meet increasingly tight environmental regulations and new fuel economy requirements has motivated system science researchers and automotive engineers to take advantage of emerging computational techniques to further advance hybrid electric vehicle and plug-in hybrid electric vehicle (PHEV) designs. In particular, research has focused on vehicle powertrain system design optimization, to reduce the fuel consumption and total energy cost while improving the vehicle's driving performance. In this work, two different natural optimization machines, namely the synchronous self-learning Pareto strategy and the elitism non-dominated sorting genetic algorithm, are implemented for component sizing of a specific power-split PHEV platform with a Toyota plug-in Prius as the baseline vehicle. To do this, a high-fidelity model of the Toyota plug-in Prius is employed for the numerical experiments using the Autonomie simulation software. Based on the simulation results, it is demonstrated that Pareto-based algorithms can successfully optimize the design parameters of the vehicle powertrain.

Supervisory Power Management Control Algorithms for Hybrid Electric Vehicles: A Survey

The growing necessity for environmentally benign hybrid propulsion systems has led to the development of advanced power management control algorithms to maximize fuel economy and minimize pollutant emissions. This paper surveys the control algorithms for hybrid electric vehicles (HEVs) and plug-in HEVs (PHEVs) that have been reported in the literature to date. The exposition ranges from parallel, series, and power split HEVs and PHEVs and includes a classification of the algorithms in terms of their implementation and the chronological order of their appearance. Remaining challenges and potential future research directions are also discussed.

Rule-Based Control Strategy With Novel Parameters Optimization Using NSGA-II for Power-Split PHEV Operation Cost Minimization

One of the major considerations in the automotive industry is the reduction of hybrid electric vehicle fuel consumption and operation cost. This paper is the first to use the nondominated sorting genetic algorithm-II (NSGA-II) for power-split plug-in hybrid electric vehicle (PHEV) applications. The NSGA-II, one of the most efficient multiobjective genetic algorithms (MOGAs), simultaneously optimized operation cost, including gasoline and electricity consumption. The Pareto optimal solutions are discussed for the parameter calibrations of the rule-based control strategy as a useful guide in PHEV development, particularly in the earlier phases. The optimized operation cost at the different power-split device (PSD) gear ratios is used to determine the ideal PSD gear ratio to further minimize the operation cost. To validate the proposed strategy, dynamic PSD and powertrain models of PHEV are developed in the numerical analysis. The two typically different driving cycles, namely, the Urban Dynamometer Driving Schedule (UDDS) and the Highway Fuel Economic Drive Schedule (HWFET), with different numbers of driving cycles, are used for control strategy optimization.

Energy management of power-split plug-in hybrid electric vehicles based on simulated annealing and Pontryagin's minimum principle

In this paper, an energy management method is proposed for a power-split plug-in hybrid electric vehicle (PHEV). Through analyzing the PHEV powertrain, a series of quadratic equations are employed to approximate the vehicle's fuel-rate, using battery current as the input. Pontryagin's Minimum Principle (PMP) is introduced to find the battery current commands by solving the Hamiltonian function. Simulated Annealing (SA) algorithm is applied to calculate the engine-on power and the maximum current coefficient. Moreover, the battery state of health (SOH) is introduced to extend the application of the proposed algorithm. Simulation results verified that the proposed algorithm can reduce fuel-consumption compared to charge-depleting (CD) and charge-sustaining (CS) mode.

Trip-Oriented Energy Management Control Strategy for Plug-In Hybrid Electric Vehicles

This paper presents a trip-oriented energy management control strategy for plug-in hybrid electric vehicle (PHEV). The proposed strategy provides system optimization and control methods to improve real-world fuel economy (FE) by optimizing the power demand distribution between fuel and battery electricity and the power delivery split between the mechanical and electrical paths in a PowerSplit PHEV architecture. A two degree of freedom system model is established to characterize the actuation dynamics and the power delivery properties of the powertrain. This paper achieves three important contributions to PHEV energy management control research: 1) the optimal control problem is solved considering both the nonlinearity of battery efficiency and the complexity of PowerSplit architecture; 2) a novel trip-oriented energy consumption preplanning method is proposed using a driving pattern-based dynamic programming approach; and 3) a feedback control system is designed to realize the optimal energy consumption process in real applications. The proposed energy management control strategy has been shown to improve FE in Ford Escape PHEVs.

Energy Management for a Power-Split Plug-in Hybrid Electric Vehicle Based on Dynamic Programming and Neural Networks

This paper focuses on building an efficient, online, and intelligent energy management controller to improve the fuel economy of a power-split plug-in hybrid electric vehicle (PHEV). Based on a detailed powertrain analysis, the battery current can be optimized to improve the fuel economy using dynamic programming (DP). Three types of drive cycles, i.e., highway, urban, and urban (congested), are classified, and six typical drive cycles are analyzed and simulated to study all the driving conditions. The online intelligent energy management controller is built, which consists of two neural network (NN) modules that are trained based on the optimized results obtained by DP methods, considering the trip length and duration. Based on whether the trip length and duration are known or unknown, the controller will choose the corresponding NN module to output the effective battery current commands to realize the energy management. Numerical simulation shows that the proposed controller can improve the fuel economy of the vehicle.