Data-driven reinforcement-learning-based hierarchical energy management strategy for fuel cell/battery/ultracapacitor hybrid electric vehicles

Sun, Haochen; Fu, Zhumu; Tao, Fazhan; Zhu, Longlong; Si, Pengju

doi:10.1016/j.jpowsour.2020.227964

Cited by 143 publications

(54 citation statements)

References 36 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The reward is the quantification of the relationship between the battery SOC and its average value. In (Sun et al, 2020), reinforcement learning (based on a Q-learning algorithm) based the hierarchical energy management strategy is applied for fuel cell hybrid electric vehicles (containing fuel cells, batteries, and supercapacitors) for lowing the computational cost and optimizing the fuel cell efficiency and the economy of energy consumption. The state variables in the RL include the SOC of the battery and the fuel cell, the power demand, and the voltage.…”

Section: Application Of Machine Learning For the Battery Energy Storage Systemmentioning

confidence: 99%

Machine learning toward advanced energy storage devices and systems

Gao

2021

iScience

115

View full text Add to dashboard Cite

Technology advancement demands energy storage devices (ESD) and systems (ESS) with better performance, longer life, higher reliability, and smarter management strategy. Designing such systems involve a trade-off among a large set of parameters, whereas advanced control strategies need to rely on the instantaneous status of many indicators. Machine learning can dramatically accelerate calculations, capture complex mechanisms to improve the prediction accuracy, and make optimized decisions based on comprehensive status information. The computational efficiency makes it applicable for real-time management. This paper reviews recent progresses in this emerging area, especially new concepts, approaches, and applications of machine learning technologies for commonly used energy storage devices (including batteries, capacitors/supercapacitors, fuel cells, other ESDs) and systems (including battery ESS, hybrid ESS, grid and microgrid-containing energy storage units, pumped-storage system, thermal ESS). The perspective on future directions is also discussed.

show abstract

Section: Application Of Machine Learning For the Battery Energy Storage Systemmentioning

confidence: 99%

Machine learning toward advanced energy storage devices and systems

Gao

2021

iScience

115

View full text Add to dashboard Cite

show abstract

“…Hence the real challenge for researchers in EV lies in reducing the power consumption while carefully designing the EV for a broader operating range, i.e., distance. They have applied DL approaches towards this problem in many works [106]- [115]. Table 7 summarizes the significant literature works considered for study in this context.…”

Section: DL In Electric Vehicle Applicationsmentioning

confidence: 99%

Deep Learning for Fault Diagnostics in Bearings, Insulators, PV Panels, Power Lines, and Electric Vehicle Applications—The State-of-the-Art Approaches

et al. 2021

View full text Add to dashboard Cite

Deep learning (DL) is an exciting field of interest for many researchers and business. Due to a massive leap in DL based research, many domains like Business, science and government sectors make use of DL for various applications. This work puts forward the importance of DL and its application in a few critical electrical segments. Initially, an introduction to Artificial Intelligence (AI) and Machine Learning (ML) is presented. Then the need for DL and the popular architectures, algorithms and frameworks used are presented. A summary of different techniques used in DL is outlined, and finally, a review on the application of deep learning techniques in some popular electrical applications is presented. Five critical electrical applications, namely identification of bearing faults, hot spots on the surface of PV panels, insulator faults, an inspection of power lines and Electric vehicles have been considered for review in this work. The primary aim of this work is to present chronologically, a survey of different areas in which it applies DL along with their architectures, frameworks and techniques to provide a deeper understanding of DL for widespread use in real-time applications.

show abstract

“…A few approaches have been examined for reinforcement learning in FCEVs. In [19], Q-learning was utilized for fuel cell/battery/ultra-capacitor hybrid electric vehicles. Here, a fuzzy filter was used to improve the performance of the RL algorithm.…”

Section: Introductionmentioning

confidence: 99%

Energy Management Strategy of Fuel Cell Electric Vehicles Using Model-Based Reinforcement Learning With Data-Driven Model Update

Lee

Cha

2021

IEEE Access

View full text Add to dashboard Cite

Fuel cell electric vehicles use fuel cells as their main power source; the vehicle is driven by an electric motor, and have an electric battery as a secondary power source that stores regenerative braking energy and assists driving. To reduce the hydrogen fuel consumption by using these fuel cells and electric batteries efficiently, an energy management strategy is needed for the proper distribution of power among them. In this study, model-based reinforcement learning was utilized for energy management. For the optimal control of a fuel-cell electric vehicle, reinforcement learning is conducted using an internal vehicle powertrain model in the learning algorithm; initially, the model is completely unknown, but the model is learned with data from experiences as the learning process progresses. Then, reinforcement learning is conducted for the environment of the driving cycle profile to optimize the control policy. In this study, vehicle simulation was conducted using standard driving cycles, and the results showed that the learning process converged steadily and that the powertrain model was well learned. The simulated fuel consumption values show that the proposed algorithm reduces fuel consumption compared to the rule-based strategy by an average of 5.7%.

show abstract

Data-driven reinforcement-learning-based hierarchical energy management strategy for fuel cell/battery/ultracapacitor hybrid electric vehicles

Cited by 143 publications

References 36 publications

Machine learning toward advanced energy storage devices and systems

Machine learning toward advanced energy storage devices and systems

Deep Learning for Fault Diagnostics in Bearings, Insulators, PV Panels, Power Lines, and Electric Vehicle Applications—The State-of-the-Art Approaches

Energy Management Strategy of Fuel Cell Electric Vehicles Using Model-Based Reinforcement Learning With Data-Driven Model Update

Contact Info

Product

Resources

About