Self-supervised reinforcement learning-based energy management for a hybrid electric vehicle

Qi, Chunyang; Zhu, Yiwen; Song, Chuanxue; Cao, Jingwei; Xiao, Feng; Zhang, Xu; Xu, Zhihao

doi:10.1016/j.jpowsour.2021.230584

Cited by 35 publications

(10 citation statements)

References 25 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Chen et al designed a semi-supervised contrastive learning algorithm and transferred knowledge of the pre-trained model to the new model using distillation learning [45]. Qi et al proposed a self-supervised reinforcement learning method that was aimed at optimizing fuel efficiency in hybrid electric vehicles [46]. This method exhibits accelerated training convergence and lower fuel consumption compared with conventional strategies, allowing for a near-global optimum fuel economy under the newly proposed driving cycle.…”

Section: Contrastive Learningmentioning

confidence: 99%

Evaluation Method of Electric Vehicle Charging Station Operation Based on Contrastive Learning

Tang¹,

Cui³

et al. 2023

BDCC

View full text Add to dashboard Cite

This paper aims to address the issue of evaluating the operation of electric vehicle charging stations (EVCSs). Previous studies have commonly employed the method of constructing comprehensive evaluation systems, which greatly relies on manual experience for index selection and weight allocation. To overcome this limitation, this paper proposes an evaluation method based on natural language models for assessing the operation of charging stations. By utilizing the proposed SimCSEBERT model, this study analyzes the operational data, user charging data, and basic information of charging stations to predict the operational status and identify influential factors. Additionally, this study compared the evaluation accuracy and impact factor analysis accuracy of the baseline and the proposed model. The experimental results demonstrate that our model achieves a higher evaluation accuracy (operation evaluation accuracy = 0.9464; impact factor analysis accuracy = 0.9492) and effectively assesses the operation of EVCSs. Compared with traditional evaluation methods, this approach exhibits improved universality and a higher level of intelligence. It provides insights into the operation of EVCSs and user demands, allowing for the resolution of supply–demand contradictions that are caused by power supply constraints and the uneven distribution of charging demands. Furthermore, it offers guidance for more efficient and targeted strategies for the operation of charging stations.

show abstract

Section: Contrastive Learningmentioning

confidence: 99%

Evaluation Method of Electric Vehicle Charging Station Operation Based on Contrastive Learning

Tang¹,

Cui³

et al. 2023

BDCC

View full text Add to dashboard Cite

show abstract

“…It is an algorithm used to describe and solve the problem that agents maximize returns or achieve specific goals through Learning strategies during their interaction with the environment. Different from the traditional evolutionary algorithms, it can realize dynamic optimization and is inclusive of the error path optimization [52]. As a classic reinforcement learning algorithm with excellent decisionmaking, Q-learning is widely used in decision-making and optimization problems.…”

Section: Stage Iii: Feature Selection Based On Reinforcement Learningmentioning

confidence: 99%

A Novel Multi-Factor Three-Step Feature Selection and Deep Learning Framework for Regional GDP Prediction: Evidence from China

Yan

2022

Sustainability

View full text Add to dashboard Cite

Gross domestic product (GDP) is an important index reflecting the economic development of a region. Accurate GDP prediction of developing regions can provide technical support for sustainable urban development and economic policy formulation. In this paper, a novel multi-factor three-step feature selection and deep learning framework are proposed for regional GDP prediction. The core modeling process is mainly composed of the following three steps: In Step I, the feature crossing algorithm is used to deeply excavate hidden feature information of original datasets and fully extract key information. In Step II, BorutaRF and Q-learning algorithms analyze the deep correlation between extracted features and targets from two different perspectives and determine the features with the highest quality. In Step III, selected features are used as the input of TCN (Temporal convolutional network) to build a GDP prediction model and obtain final prediction results. Based on the experimental analysis of three datasets, the following conclusions can be drawn: (1) The proposed three-stage feature selection method effectively improves the prediction accuracy of TCN by more than 10%. (2) The proposed GDP prediction framework proposed in the paper has achieved better forecasting performance than 14 benchmark models. In addition, the MAPE values of the models are lower than 5% in all cases.

show abstract

“…During the actual slewing operation of the load, the EM will work inefficiently resulting in low energy utilization and serious energy loss. The hybrid power system consists of two or more powergenerating devices [3] , and at least one power-generating device can achieve energy storage. Among them, electrohydraulic hybrid power technology integrates EM drive technology and hydraulic power technology [4] [5] , which not only takes advantage of the EM's zero emissions and low pollution but also takes full advantage of the hydraulic power source's ability to instantly provide high torque, high power density, and fast energy storage and discharge.…”

Section: Introductionmentioning

confidence: 99%

Characterization Analysis of a New Energy Regenerative Electro-Hydraulic Hybrid Rotary System

Du,

Liu,

Quan

2024

IEEE Access

View full text Add to dashboard Cite

The rotary motion field of heavy equipment is a large energy-consuming field, characterized by frequent cyclical start-stop, significant load inertia, and braking energy is usually dissipated as thermal energy, for energy regeneration, no good solution has been found. This paper proposes a new energyregenerative hybrid drive (ERHD) system and defines three operating modes. Taking the reversible rolling mill's roll as load, and taking the energy consumption of electric motor (EM) and peak power as evaluation indexes, the influence law of key component parameters on the driving characteristics of the system is studied. Simulation models for the pure electric motor drive (PEMD) system and ERHD system were established by AMESim, along with a rule-based energy management strategy (EMS). The results show that the hydraulic pump/motor displacement has the greatest influence on braking time, peak power, and energy consumption. Under the optimal parameters, the proposed ERHD system can reduce the peak power of EM by 45.75% and energy consumption by 48.33% compared with the PEMD system. Starting and braking times were reduced by 44.54% and 1.21%, respectively. Based on the test results, the ERHD system reduces start-up time by 22.70% and peak power by 23.72%. The proposed system will significantly contribute to developing the rotary motion field of heavy equipment in terms of energy efficiency, low carbon and green.

show abstract

Self-supervised reinforcement learning-based energy management for a hybrid electric vehicle

Cited by 35 publications

References 25 publications

Evaluation Method of Electric Vehicle Charging Station Operation Based on Contrastive Learning

Evaluation Method of Electric Vehicle Charging Station Operation Based on Contrastive Learning

A Novel Multi-Factor Three-Step Feature Selection and Deep Learning Framework for Regional GDP Prediction: Evidence from China

Characterization Analysis of a New Energy Regenerative Electro-Hydraulic Hybrid Rotary System

Contact Info

Product

Resources

About