Article
Thermodynamics
Chunyang Qi, Yiwen Zhu, Chuanxue Song, Guangfu Yan, Da Wang, Feng Xiao, Xu Zhang, Jingwei Cao, Shixin Song
Summary: This research introduces a novel reinforcement learning-based deep Q-learning algorithm for the energy management strategy of HEVs. The proposed method not only addresses the issue of sparse reward during training, but also achieves optimal power distribution. Additionally, the hierarchical structure of the algorithm enhances exploration of the vehicle environment, leading to improved training efficiency and reduced fuel consumption.
Article
Energy & Fuels
Hailong Zhang, Jiankun Peng, Hanxuan Dong, Huachun Tan, Fan Ding
Summary: The economy-oriented automated hybrid electric vehicles (HEV) have the potential to save energy by optimizing driving behaviors and power distribution. Recent advances in the ecological car following issue of HEV focus on integrating adaptive cruise control (ACC) and energy management system (EMS) for collaborative optimization. However, the current control frameworks have limitations in optimization. To address this, a hierarchical reinforcement learning based ACC-EMS strategy is proposed, which significantly improves training speed and stability in car-following scenarios.
Article
Energy & Fuels
Haroune Aouzellag, Bessam Amrouche, Koussaila Iffouzar, Djamal Aouzellag
Summary: This paper proposes an energy management strategy based on hysteresis control, using the expression of the energy storage system converter efficiency as the sole input variable. Simulation results show that the strategy allows energy distribution and reduces energy consumption.
JOURNAL OF ENERGY STORAGE
(2022)
Article
Chemistry, Physical
Meiling Yue, Zeina Al Masry, Samir Jemei, Noureddine Zerhouni
Summary: This paper presents a health management strategy for fuel cell hybrid electric vehicles based on online adaptive prognostics technology, which effectively improves the durability of on-board fuel cells through online health monitoring and decision fusion methods.
INTERNATIONAL JOURNAL OF HYDROGEN ENERGY
(2021)
Article
Thermodynamics
Razieh Ghaderi, Mohsen Kandidayeni, Loic Boulon, Joao P. Trovao
Summary: This paper proposes a three-layer online EMS for efficient power distribution in a multi-stack fuel cell hybrid electric vehicle. The method continuously updates the characteristics of each fuel cell and battery using online estimators, and uses a rule-based approach to improve the calculation speed. The power distribution is achieved using a Q-learning algorithm based on reinforcement learning. The proposed method significantly reduces the trip cost compared to other online strategies and has a slightly higher cost compared to the offline strategy.
ENERGY CONVERSION AND MANAGEMENT
(2023)
Article
Thermodynamics
Chunyang Qi, Chuanxue Song, Feng Xiao, Shixin Song
Summary: This paper investigates the generalization capability of energy management strategies for hybrid electric vehicles and proposes a multi-agent reinforcement learning algorithm. By analyzing typical features and using an auxiliary agent, the generalization performance of energy management strategies is improved.
Article
Energy & Fuels
Hao Zhang, Qinhao Fan, Shang Liu, Shengbo Eben Li, Jin Huang, Zhi Wang
Summary: This study aims to investigate the optimal control strategy for dedicated hybrid engine-driven plug-in hybrid electric vehicles under real driving conditions. The hierarchical energy management strategy proposed in the study shows superior performance in terms of fuel economy and NOx emissions.
Article
Engineering, Electrical & Electronic
Yue Hu, Hui Xu, Zhonglin Jiang, Xinyu Zheng, Jianfeng Zhang, Wenhui Fan, Kun Deng, Kun Xu
Summary: This paper proposes a supplementary learning controller (SLC) based on deep reinforcement learning (DRL) to compensate for an existing rule-based energy management system (EMS) for hybrid electric vehicles (HEVs). The SLC works alongside the rule-based EMS and reduces the uncertainty of the algorithm to the system. A distributed architecture is designed for the DRL-based SLC, where each vehicle's SLC interacts with its own driving cycle, shares a neural network, and sends experience data to the cloud for learning updates.
IEEE TRANSACTIONS ON VEHICULAR TECHNOLOGY
(2023)
Article
Engineering, Electrical & Electronic
Zhumu Fu, Haocong Wang, Fazhan Tao, Baofeng Ji, Yongsheng Dong, Shuzhong Song
Summary: In this paper, an energy management strategy based on a hierarchical power splitting structure and deep reinforcement learning is proposed to address the challenges in energy management for fuel cell hybrid electric vehicles equipped with battery and ultracapacitor. The strategy optimizes power allocation and improves working efficiency and fuel economy through adaptive filtering and heuristic techniques.
IEEE TRANSACTIONS ON VEHICULAR TECHNOLOGY
(2022)
Article
Engineering, Electrical & Electronic
Bo Hu, Jiaxi Li
Summary: With the development of artificial intelligence and machine learning, reinforcement learning has opened up new possibilities for hybrid electric vehicle energy management. However, current issues limit its application in industrial energy management strategy tasks. To overcome this, an adaptive hierarchical energy management strategy combining heuristic equivalent consumption minimization strategy knowledge and deep deterministic policy gradient algorithm is proposed. Experimental results show that the proposed strategy outperforms other benchmark strategies in terms of fuel consumption.
IEEE TRANSACTIONS ON TRANSPORTATION ELECTRIFICATION
(2022)
Article
Thermodynamics
Jianhao Zhou, Siwu Xue, Yuan Xue, Yuhui Liao, Jun Liu, Wanzhong Zhao
Summary: In this study, a deep reinforcement learning algorithm, TD3, is used to develop an intelligent energy management strategy (EMS) for hybrid electric vehicles, including a local controller (LC) and a hybrid experience replay method (HER). The improved TD3-based EMS shows the best fuel optimization performance, fastest convergence speed, and highest robustness under different driving cycles.
Review
Green & Sustainable Science & Technology
M. Cha, H. Enshaei, H. Nguyen, S. G. Jayasinghe
Summary: This paper provides a comprehensive review of various power management strategies (PMSs) for fuel cell hybrid electric vehicles (FCHEVs), with a focus on data acquisition and management. The classification and performance of PMSs are analyzed to determine their objectives and characteristics for optimization. Furthermore, the application of PMSs in the marine industry, particularly in ship propulsion systems, is examined to identify research trends and challenges. The importance of considering component sizing, control, and health-conscious management for optimizing cost-effectiveness and the life cycle of FC and battery electric ferry applications is also highlighted.
RENEWABLE & SUSTAINABLE ENERGY REVIEWS
(2023)
Article
Thermodynamics
Yang Zhou, Alexandre Ravey, Marie-Cecile Pera
Summary: In this paper, a real-time cost minimization energy management strategy for fuel cell/battery-based hybrid electric vehicles is proposed using model predictive control. Results show that the strategy can effectively reduce operating costs and extend fuel cell lifetime, demonstrating good real-time practicality.
ENERGY CONVERSION AND MANAGEMENT
(2021)
Article
Engineering, Electrical & Electronic
Mohammadreza Moghadari, Mohsen Kandidayeni, Loic Boulon, Hicham Chaoui
Summary: This paper compares the operating cost of a single-stack and a multi-stack fuel cell hybrid electric vehicle (FC-HEV), including hydrogen consumption and degradation of the fuel cell (FC). A hierarchical energy management strategy (EMS) is developed for the multi-stack system. Results show that the FC-HEV with a multi-stack structure has lower hydrogen and degradation costs compared to the single-stack structure.
IEEE TRANSACTIONS ON VEHICULAR TECHNOLOGY
(2023)
Article
Thermodynamics
Wei Cui, Naxin Cui, Tao Li, Zhongrui Cui, Yi Du, Chenghui Zhang
Summary: This study proposes a multi-objective hierarchical energy management strategy to improve the performance of connected plug-in hybrid electric vehicles (PHEVs). The strategy incorporates resistance network-triggered motion planning and convex torque optimization based on the alternating direction method of multipliers (ADMM), aiming to comprehensively optimize energy saving, safety, traffic efficiency, and computational efficiency.
Article
Engineering, Electrical & Electronic
Qiao Xue, Junqiu Li, Zheng Chen, Yuanjian Zhang, Yonggang Liu, Jiangwei Shen
Summary: This paper proposes a deep learning method for online capacity estimation of lithium-ion batteries. The proposed algorithm utilizes a predictive model that combines a convolutional neural network and a long short-term memory unit for automatic feature extraction and target estimation. Experimental results demonstrate that this method can accurately estimate battery capacity using only short voltage and current data.
IEEE TRANSACTIONS ON VEHICULAR TECHNOLOGY
(2023)
Article
Multidisciplinary Sciences
Chunyan Shuai, Fang Yang, Wencong Wang, Jun Shan, Zheng Chen, Xin Ouyang, Vickie E. Baracos, Marilia Cravo
Summary: By analyzing 181,282 charge records, researchers have developed a deep neural network algorithm using IoT technology that can automatically capture charge feature variables, determine their dependencies, and identify abnormal charge behaviors, effectively ensuring the charging safety of over 20 million E-bicycles.
Article
Thermodynamics
Yuanjian Zhang, Bingzhao Gao, Jingjing Jiang, Chengyuan Liu, Dezong Zhao, Quan Zhou, Zheng Chen, Zhenzhen Lei
Summary: In this paper, a cooperative power management strategy is proposed for range extended electric vehicles (REEVs), utilizing vehicle-environment cooperation and abundant information exchanged in the internet of vehicles (IoVs). The strategy integrates the self-learning explicit equivalent minimization consumption strategy (SL-eECMS) and adaptive neuro-fuzzy inference system (ANFIS) based online charging management within on-board power sources in the REEV. Simulation results and hardware-in-the-loop (HIL) test demonstrate the effectiveness and efficiency of the proposed strategy in managing power flow within power sources in the REEV.
Article
Thermodynamics
Hongqian Zhao, Zheng Chen, Xing Shu, Jiangwei Shen, Yonggang Liu, Yuanjian Zhang
Summary: Accurate and early detection of voltage faults is crucial for protecting property and passengers. This study develops a precise voltage prediction and fault diagnosis method using a gated recurrent unit neural network and incremental training. The method can predict battery voltage in advance and detect faults with high accuracy.
Article
Engineering, Electrical & Electronic
Jie Li, Yonggang Liu, Abbas Fotouhi, Xiangyu Wang, Zheng Chen, Yuanjian Zhang, Liang Li
Summary: In this research, a learning-based method is used to achieve satisfactory fuel economy for connected plug-in hybrid electric vehicles (PHEVs) in car-following scenarios. By leveraging a data-driven energy consumption model and considering the nonlinear efficiency characteristics, an advanced ADP scheme is designed for connected PHEVs. The cooperative information is also incorporated to improve fuel economy and driving safety.
IEEE TRANSACTIONS ON VEHICULAR TECHNOLOGY
(2023)
Article
Engineering, Electrical & Electronic
Simin Wu, Zheng Chen, Shiquan Shen, Jiangwei Shen, Fengxiang Guo, Yonggang Liu, Yuanjian Zhang
Summary: In this study, a hierarchical cooperative eco-driving control strategy is proposed, enabling vehicles in a platoon to plan ecological speed trajectories and pass intersections smoothly via V2X communication. Unlike existing eco-driving studies, this research focuses on the cooperative optimization of a group of vehicles at consecutive signalized intersections. Simulation results show that the proposed strategy can improve vehicle-following behaviors and platoon performance, with a 26.10% reduction in overall energy consumption and a 2.83% reduction in trip time compared to manual driving.
IET INTELLIGENT TRANSPORT SYSTEMS
(2023)
Article
Thermodynamics
Jie Li, Abbas Fotouhi, Wenjun Pan, Yonggang Liu, Yuanjian Zhang, Zheng Chen
Summary: This study proposes an eco-driving approach with a hierarchical framework to be leveraged at signalized intersections considering the impact of traffic uncertainty. The proposed method utilizes a queue-based traffic model to estimate traffic uncertainty and generate dynamic modified traffic light information. Additionally, a deep reinforcement learning-based controller is constructed to optimize velocity, reducing energy consumption and ensuring driving safety. Simulation results demonstrate the effectiveness of the proposed control strategy in improving energy economy and preventing unnecessary idling in uncertain traffic scenarios, as compared to approaches that ignore traffic uncertainty. The proposed method is also adaptable to different traffic scenarios and exhibits energy efficiency.
Article
Thermodynamics
Jiangwei Shen, Wensai Ma, Xing Shu, Shiquan Shen, Zheng Chen, Yonggang Liu
Summary: This article proposes a method based on charging voltage prediction and machine learning for reliable estimation of lithium-ion battery state of health (SOH). Correlated feature variables are identified by analyzing the raw charging voltage distribution, and a polynomial-based prediction model is used to estimate a wide range of charging voltage. The extreme learning machine algorithm is then employed for online SOH estimation. Experimental results demonstrate that the proposed method can provide reliable SOH estimation with an error of less than 2.02% in short-term random charging scenarios.
Article
Thermodynamics
Yonggang Liu, Qianyou Chen, Jie Li, Yuanjian Zhang, Zheng Chen, Zhenzhen Lei
Summary: This study investigates the optimization problem of eco-routing on a road network for heterogeneous continuous vehicle flow, and proposes a collaborative heterogeneous multi-vehicle eco-routing optimization strategy to improve the overall economy in the road network.
Article
Automation & Control Systems
Chunyan Shuai, Yu Sun, Xiaoqi Zhang, Fang Yang, Xin Ouyang, Zheng Chen
Summary: The widespread use of electric bicycles (E-bikes) has raised concerns about charging safety. However, diagnosing charging safety for E-bikes online is challenging due to limited data and multiple factors involved. This article proposes a nonintrusive intelligent diagnosis scheme on the inputted power grid side to overcome this challenge. By using an improved dynamic time warping model, the proposed scheme can accurately identify abnormal charging processes and achieve a high precision rate of 94%.
IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS
(2023)
Article
Multidisciplinary Sciences
Jingyang Li, Fengxiang Guo, Wei Li, Bijiang Tian, Zheng Chen, Sirou Qu
Summary: Considerable evidence suggests that the decline in physiological abilities of older drivers leads to a reduction in the visual and psychomotor functions required for safe driving. This study recruited both older and younger drivers for a driving simulation experiment and compared their driving behaviors, establishing driving behavior graphs. The results indicate that older drivers have slower responses in observing traffic information and applying brakes and steering, providing valuable insights for analyzing driving behavior and safety for older individuals.
Article
Engineering, Electrical & Electronic
Shuai Liu, Fan Ren, Ping Li, Zhijie Li, Hao Lv, Yonggang Liu
Summary: This paper presents a novel approach for automatically identifying test scenarios for autonomous driving through deep unsupervised learning. The proposed unsupervised framework combines isolation forest algorithm, one-dimensional residual convolutional autoencoder, and information entropy-optimized K-means algorithm to automatically identify typical and extreme scenarios from NDD.
WORLD ELECTRIC VEHICLE JOURNAL
(2023)
Article
Engineering, Electrical & Electronic
Guo Yang, Shihuan Liu, Ming Ye, Chengcheng Tang, Yi Fan, Yonggang Liu
Summary: The ramp merging decision is crucial in ensuring the efficiency and safety of the entire merging process. However, due to the unavoidable nature of on-ramp merging, limitations of the road environment, and conflicts between merging and following vehicles, it is difficult for human drivers to make optimal decisions in complex merging scenarios.
WORLD ELECTRIC VEHICLE JOURNAL
(2023)
Article
Multidisciplinary Sciences
Chunyan Shuai, Fang Yang, Wencong Wang, Jun Shan, Zheng Chen, Xin Ouyang
Summary: The widespread use of electric bicycles has led to numerous charging accidents. However, diagnosing charging faults online is challenging due to the lack of standard chargers, inconsistent communication methods, and limited access to battery status. The development of the Internet of Things allows for the collection of charger input current information on a cloud platform, providing an alternative approach to identify underlying charging abnormalities. By analyzing 181,282 charge records, a deep neural network algorithm has been developed to automatically capture charge feature variables, determine their dependencies, and identify abnormal charge behaviors. With an average accuracy of 85%, the algorithm effectively diagnoses charging faults, ensuring the safety of over 20 million E-bicycles after extensive validation. Furthermore, this diagnostic framework can be extended to real-time charge safety detection for electric vehicles and similar vehicles.
Review
Green & Sustainable Science & Technology
Jie Li, Abbas Fotouhi, Yonggang Liu, Yuanjian Zhang, Zheng Chen
Summary: With the development of communication and automation technologies, the energy-saving potential of connected and automated vehicles (CAVs) has been highlighted. This study systematically summarizes the state-of-the-art in eco-driving, which is the automatic planning of ecological driving behaviors to reduce energy consumption. The study discusses the basic principles of eco-driving, classifies related studies, and emphasizes the potential for cooperative eco-driving in terms of energy saving. The study provides potential development trends for eco-driving techniques.
RENEWABLE & SUSTAINABLE ENERGY REVIEWS
(2024)
Article
Green & Sustainable Science & Technology
Lars odegaard Bentsen, Narada Dilp Warakagoda, Roy Stenbro, Paal Engelstad
Summary: This study investigates uncertainty modeling in wind power forecasting using different parametric and non-parametric methods. Johnson's SU distribution is found to outperform Gaussian distributions in predicting wind power. This research contributes to the literature by introducing Johnson's SU distribution as a candidate for probabilistic wind forecasting.
JOURNAL OF CLEANER PRODUCTION
(2024)
Article
Green & Sustainable Science & Technology
Xing Liu, Qiuchen Wang, Yunhao Wen, Long Li, Xinfang Zhang, Yi Wang
Summary: This study analyzes the characteristics of process parameters in three lean gas ethane recovery processes and establishes a prediction and multiobjective optimization model for ethane recovery and system energy consumption. A new method for comparing ethane recovery processes for lean gas is proposed, and the addition of extra coolers improves the ethane recovery. The support vector regression model based on grey wolf optimization demonstrates the highest prediction accuracy, and the multiobjective multiverse optimization algorithm shows the best optimization performance and diversity in the solutions.
JOURNAL OF CLEANER PRODUCTION
(2024)
Article
Green & Sustainable Science & Technology
Cairong Song, Haidong Yang, Xian-Bing Meng, Pan Yang, Jianyang Cai, Hao Bao, Kangkang Xu
Summary: The paper proposes a novel deep learning-based prediction framework, aTCN-LSTM, for accurate cooling load predictions. The framework utilizes a gate-controlled multi-head temporal convolutional network and a sparse probabilistic self-attention mechanism with a bidirectional long short-term memory network to capture both temporal and long-term dependencies in the cooling load sequences. Experimental results demonstrate the effectiveness and superiority of the proposed method, which can serve as an effective guide for HVAC chiller scheduling and demand management initiatives.
JOURNAL OF CLEANER PRODUCTION
(2024)
Article
Green & Sustainable Science & Technology
Zhe Chen, Xiaojing Li, Xianli Xia, Jizhou Zhang
Summary: This study uses survey data from the Loess Plateau in China to evaluate the impact of social interaction on the adoption of soil and water conservation (SWC) technology by farmers. The study finds that social interaction increases the likelihood of farmers adopting SWC, and internet use moderates this effect. The positive impact of social interaction on SWC adoption is more pronounced for farmers in larger villages and those who join cooperative societies.
JOURNAL OF CLEANER PRODUCTION
(2024)
Article
Green & Sustainable Science & Technology
Chenghua Zhang, Yunfei Yan, Kaiming Shen, Zongguo Xue, Jingxiang You, Yonghong Wu, Ziqiang He
Summary: This paper reports a novel method that significantly improves combustion performance, including heat transfer enhancement under steady-state conditions and adaptive stable flame regulation under velocity sudden increase.
JOURNAL OF CLEANER PRODUCTION
(2024)
