Research on Energy Management Strategies for Fuel Cell Hybrid Vehicles Based on Time Classification

Lihua Ye, Zixing Zhang, Qinglong Zhao, Zhao Xu, He Zhou, A. ShiEnergies 2025 · DOI

TL;DR

This paper introduces a novel time-classification-based energy management strategy for fuel cell hybrid vehicles (FCHVs) that significantly reduces hydrogen consumption compared to traditional condition-identification methods. By leveraging temporal information alongside driving conditions, the strategy achieves a 2.707% reduction in equivalent hydrogen consumption, enhancing overall energy efficiency.

What Problem Does This Paper Address?

The transportation sector is a major contributor to CO2 emissions, driving the need for more efficient and environmentally friendly vehicles. Fuel cell hybrid vehicles (FCHVs) offer a promising solution, but optimizing their energy consumption is crucial. Existing energy management strategies (EMS) often struggle with accurately identifying driving conditions in real-time, leading to suboptimal power allocation and increased energy usage. This paper aims to address the limitations of current condition-identification EMS by proposing a time-classification-based approach that improves accuracy and reduces hydrogen consumption.

Methods

The study establishes a mathematical model for the FCHV powertrain, including models for the fuel cell, battery, and motor. Driving conditions are constructed using actual collected vehicle data. The core of the proposed method involves:

  • Time Classification: Dividing driving data into distinct time periods (e.g., morning rush hour, mid-morning) and associating specific characteristics with these periods.
  • Working Condition Identification: Employing a combination of Principal Component Analysis (PCA) for dimensionality reduction and K-means clustering to categorize driving segments into urban, suburban, and high-speed conditions. A Support Vector Machine (SVM) is then trained for online condition identification.
  • Pontryagin's Minimum Principle (PMP): This optimization principle is used to determine the optimal power allocation for minimizing equivalent hydrogen consumption. The study iteratively solves for the optimal covariate ($\lambda$) using a target-hitting method.
  • Time-Based EMS: The proposed strategy utilizes the time-classification information to inform the PMP optimization, allowing for more precise covariate selection and thus more efficient power allocation compared to strategies relying solely on identified working conditions.

Datasets

The study utilizes actual collected vehicle driving data for training and simulation. Specific details on the dataset's origin or size are not explicitly provided beyond it being collected over an entire year.

Key Insights and Ideas

  • Temporal Context Matters: Simply identifying a driving condition (e.g., "urban") isn't enough; when that condition occurs (e.g., "urban during rush hour") provides crucial information for optimizing energy management.
  • PMP's Adaptability: Pontryagin's Minimum Principle, when combined with time-classification, proves to be a powerful tool for dynamic power allocation, effectively balancing the needs of the fuel cell and battery.
  • Improved Accuracy, Reduced Consumption: The time-classification approach leads to more accurate identification of operating scenarios, which directly translates to lower hydrogen consumption.

Key Results

  • The proposed time-classification-based EMS reduced equivalent hydrogen consumption by 2.707% compared to the condition-identification strategy.
  • The condition identification strategy achieved a validation accuracy of 97.97% and a test accuracy of 97.30%.
  • The time-based classification strategy resulted in a 3.788% reduction in equivalent hydrogen consumption over 10 simulation cycles compared to the condition-based identification strategy.
  • Statistical analysis confirmed a significant difference between the two strategies, with the time-classification strategy showing a large effect size (Cohen's d = 1.04) and lower hydrogen consumption.

Limitations and Future Work

The authors acknowledge several limitations:

  • Ignored Dynamic Aging: The study did not account for the dynamic aging characteristics of the fuel cell stack, assuming constant performance.
  • Environmental Factors: The influence of ambient temperature on the energy management system and thermal management was not included.
  • Lack of Real-World Testing: The energy management strategy was validated through simulation only and has not been applied to real vehicle testing.

Future work could explore incorporating these factors to further enhance the EMS. The study also suggests that further research into the relationship between hydrogen consumption, lifetime shortening, and performance degradation is needed.

Summary updated

Abstract

In order to minimize the carbon emission and energy consumption of fuel cell hybrid vehicles and, at the same time, solve the problem of low accuracy of working condition identification in the working condition identification strategy, this paper proposes an energy management strategy for SUVs on the basis of the working condition identification energy management strategy by using the time classification method. First, the mathematical model of the whole vehicle power system is established, and the driving conditions are constructed using actual collected vehicle driving data. On this basis, the working condition identification model was established, and then the energy management strategy of time working condition classification was established on the basis of the working condition identification model, and the equivalent hydrogen consumption of the two strategies was calculated by the Pontryagin minimization strategy. The results show that the strategy proposed in this paper reduces the equivalent hydrogen consumption by 2.707% compared with the condition identification strategy. This improvement not only greatly improves the energy efficiency of the fuel cell hybrid vehicle but also provides new ideas for the optimization of future energy management strategies.
BibTeX
@inproceedings{ye2025research,
  author = {Lihua Ye and Zixing Zhang and Qinglong Zhao and Zhao Xu and He Zhou and A. SHI},
  title = {{Research on Energy Management Strategies for Fuel Cell Hybrid Vehicles Based on Time Classification}},
  booktitle = {Energies},
  year = {2025},
  doi = {10.3390/en18082103},
}
Figure