Round 1

Reviewer 1 Report

Comments and Suggestions for Authors This paper reviews the current status of research on fuel cell hybrid power systems for commercial vehicles (buses, heavy trucks), and systematically compares three types of hybrid configurations: fuel cell-battery, fuel cell-supercapacitor, and fuel cell-battery-supercapacitor, and their applications in passenger cars, buses, and trucks. The paper focuses on analyzing the energy management strategies of indirect connection structures (such as predictive control, fuzzy control, adaptive sliding mode control, etc.), aiming to improve the economy and durability of the system. The manuscript classification system is comprehensive, covering mainstream hybrid configurations and application scenarios, and summarizes various energy management strategies, with both academic and engineering reference values. However, the manuscript still has certain shortcomings, such as: lack of actual application data to support the conclusions; insufficient discussion of the technical details of key actuators, such as drive motors; and the breadth of literature coverage needs to be strengthened.   1. The literature categorizes fuel cell hybrid power systems into three types (fuel cell battery, fuel cell supercapacitor, fuel cell battery supercapacitor) and discusses them separately for passenger cars, buses, and heavy-duty trucks. However, the classification criteria are not clearly stated (such as whether they are based on power demand or application scenarios). Suggest supplementing the theoretical or empirical basis for classification criteria, such as whether to consider the impact of typical operating conditions of different vehicle models (such as urban circulation vs. long-distance transportation) on system selection.   2. Page 2, 'However, fuel cells also have some drawbacks, such as slow cold start, poor dynamic characteristics, and the inability of braking energy recovery [9]. Reasonable matching of fuel cell vehicle power system parameters and energy management strategies can make the key components of the power system work in coordination, and improve the economic performance of the vehicle under the premise of satisfying its dynamics [10].'     The primary drawback of the fuel cell systems is their limited lifespan and performance degradation, which is not described at all. Fuel cells will experience a gradual decline in performance during long-term operation. This performance degradation is caused by a variety of complex factors, including the degradation of electrode materials, a loss of catalysts, mechanical damage to the membrane electrode assembly, and fluctuations in operating conditions. Performance degradation not only affects the efficiency and output power of fuel cells but also shortens their service life and increases maintenance and replacement costs (10.3390/en17123050).      Over prolonged operation periods, the performance of PEMFC gradually deteriorates due to intricate operational conditions and component aging, eventually reaching the minimum acceptable threshold. Prognosis and Health Management can enable proactive interventions to prevent fuel cell failures, consequently prolonging the life of PEMFC, see fuel cell life prediction considering the recovery phenomenon of reversible voltage loss. This content should be further supplied in the main text.    3. The literature reviews various strategies (such as fuzzy control, model predictive control, reinforcement learning, etc.), but lacks a systematic comparison of the advantages and disadvantages of these strategies. The differences in key indicators such as hydrogen consumption, battery life, and real-time performance between different strategies have not been quantitatively analyzed; The complexity of the strategy and the feasibility of actual engineering implementation (such as computing resource requirements) have not been discussed   4. The literature mentions issues such as slow cold start and poor dynamic response of fuel cells, but does not fully explore how existing research can address these challenges. Have the measures to suppress the decay of fuel cell lifespan been validated over the long term? Moreover, the health state estimation and long-term durability prediction for vehicular PEM fuel cell stacks under dynamic operational conditions are rather important, and related descriptions should be further added.   5. The literature focuses on technical performance, but does not evaluate the cost-effectiveness of different hybrid systems. Is the high cost of fuel cell supercapacitor systems offset by their extended lifespan? Does the current situation of hydrogen fuel infrastructure limit the application of heavy trucks?   6. The direction proposed in the literature for "technological breakthroughs, system optimisation, application expansion, and policy support" is relatively general. Suggestion refinement: Policy level: Which countries' subsidy models can be used for reference

Author Response

Thank you very much for taking the time to review this manuscript. Please find the detailed responses below. The manuscript was further revised in terms of language and optimized for image typography, proportions, and so on.

Comments 1: The literature categorizes fuel cell hybrid power systems into three types (fuel cell battery, fuel cell supercapacitor, fuel cell battery supercapacitor) and discusses them separately for passenger cars, buses, and heavy-duty trucks. However, the classification criteria are not clearly stated (such as whether they are based on power demand or application scenarios). Suggest supplementing the theoretical or empirical basis for classification criteria, such as whether to consider the impact of typical operating conditions of different vehicle models (such as urban circulation vs. long-distance transportation) on system selection.

Response 1: Thank you for pointing this out. We agree with your suggestion. We have clarified the rationale for categorizing fuel cell hybrid systems by topology and by vehicle type, and have made the appropriate changes. Location of the change - page 4, paragraph 2, line 148, highlighted in yellow.

 

Comments 2: Page 2, 'However, fuel cells also have some drawbacks, such as slow cold start, poor dynamic characteristics, and the inability of braking energy recovery [9]. Reasonable matching of fuel cell vehicle power system parameters and energy management strategies can make the key components of the power system work in coordination, and improve the economic performance of the vehicle under the premise of satisfying its dynamics [10].' The primary drawback of the fuel cell systems is their limited lifespan and performance degradation, which is not described at all. Fuel cells will experience a gradual decline in performance during long-term operation. This performance degradation is caused by a variety of complex factors, including the degradation of electrode materials, a loss of catalysts, mechanical damage to the membrane electrode assembly, and fluctuations in operating conditions. Performance degradation not only affects the efficiency and output power of fuel cells but also shortens their service life and increases maintenance and replacement costs (10.3390/en17123050). Over prolonged operation periods, the performance of PEMFC gradually deteriorates due to intricate operational conditions and component aging, eventually reaching the minimum acceptable threshold. Prognosis and Health Management can enable proactive interventions to prevent fuel cell failures, consequently prolonging the life of PEMFC, see fuel cell life prediction considering the recovery phenomenon of reversible voltage loss. This content should be further supplied in the main text.

Response 2: Thank you for your professional correction on the issue of fuel cell performance degradation and lifetime. The mechanism of performance degradation in long-term operation and the health management strategy you mentioned are of great significance to improve the research of this paper. According to your suggestions, we have made corresponding changes, the specific location of the changes - page 2, paragraph 1, line 48, yellow highlighting.

 

Comments 3: The literature reviews various strategies (such as fuzzy control, model predictive control, reinforcement learning, etc.), but lacks a systematic comparison of the advantages and disadvantages of these strategies. The differences in key indicators such as hydrogen consumption, battery life, and real-time performance between different strategies have not been quantitatively analyzed; The complexity of the strategy and the feasibility of actual engineering implementation (such as computing resource requirements) have not been discussed.

Response 3: Your in-depth review of the article is often appreciated. In fact, the advantages and disadvantages of various control strategies, the differences in key indexes and engineering feasibility have been discussed in the paper. In terms of advantages and disadvantages, the characteristics of fuzzy control and model predictive control in terms of parameter tuning, dynamic optimization, and computation volume are described through specific topology and working condition analyses, e.g., in the chapters of passenger cars and buses; in terms of key indexes, in the sections of optimization of hydrogen consumption and analysis of module life, the effects of different strategies are compared with real cases, e.g., the reinforcement learning strategy reduces hydrogen consumption, the MPC strategy prolongs fuel cell life, and the engineering feasibility is discussed in chapter 5.2 in terms of light weighting and fuel cell life. In chapter 5.2, the engineering feasibility section explains the requirements of different strategies on computational resources from the perspective of light weighting and structural integration design. Although these contents are not presented in the form of centralized comparisons, they are all throughout the discussion in each chapter and are closely integrated with the core of the research.

 

Comments 4: The literature mentions issues such as slow cold start and poor dynamic response of fuel cells, but does not fully explore how existing research can address these challenges. Have the measures to suppress the decay of fuel cell lifespan been validated over the long term? Moreover, the health state estimation and long-term durability prediction for vehicular PEM fuel cell stacks under dynamic operational conditions are rather important, and related descriptions should be further added.

Response 4: Thank you for your valuable inputs. As a review focusing on fuel cell hybrid system architecture and energy management strategies, this paper explains how to avoid the inherent defects of fuel cells in a synergistic way through systematic analysis of hybrid design and adaptive control strategies (e.g., ASTSMC, fuzzy control) in the optimization of cold-start and dynamic response. ) have not emphasized long-term measurement data, but have demonstrated feasible paths to inhibit aging based on simulation and short-term validation (e.g., hardware-in-the-loop), and have acknowledged that long-term validation requires continuous collaboration between industry, academia, and research. Regarding the prediction of the health state, although this paper focuses on comparing the system-level strategies without in-depth investigation into the component-level algorithms, it mentions SOH self-adaptive mechanism in the degradation model of the fuel cell and the optimization of the game (Ref. [76]), and calls for the integration of digital control into the system in the outlook. And in the outlook, we call for complementing that direction with digital twin technology to echo your concerns about durability refinement management.

 

Comments 5: The literature focuses on technical performance, but does not evaluate the cost-effectiveness of different hybrid systems. Is the high cost of fuel cell supercapacitor systems offset by their extended lifespan? Does the current situation of hydrogen fuel infrastructure limit the application of heavy trucks?

Response 5: Thank you for your valuable comments. Although this paper focuses on the technical performance review, the correlation between cost-effectiveness and infrastructure has been implicitly analyzed in the key chapters: Section 5.1 points out the core contradiction of fuel cell cost (platinum-based catalysts account for more than 40%) and puts forward the path of non-precious-metal catalysts to reduce the cost (the cost can be reduced to 1/20); Section 3.3 shows that high initial cost can be offset by life compensation through the evidence of supercapacitors “absorbing peak currents to extend the life of battery Section 3.3 demonstrates that the high initial cost can be offset by lifetime compensation through supercapacitors “absorbing peak current to extend battery life by 300%” and Section 4.1 demonstrates that the three-source system “reduces performance degradation by 85.4%”, and the 5.9% reduction in hydrogen consumption by the half-power prediction strategy in Table 1 hints at the potential for optimizing operating costs. The hydrogen infrastructure constraints that you so sensitively point out have been initially touched upon in Section 5.3 on range enhancement of heavy trucks (e.g., Daimler GenH2) and Section 5.4 on policy support frameworks (e.g., the European Union's CBAM mechanism), and we fully agree that the complexity of this issue is a direction that future research needs to explore in conjunction with industry, for example, by combining commercial models to Quantifying the correlation between hydrogen refueling network coverage and TCO. While this paper is a technical performance review and does not go into the economic analysis of the infrastructure, the perspectives you have provided will provide key guidance for the subsequent work. Thank you again for your valuable input.

 

Comments 6: The direction proposed in the literature for "technological breakthroughs, system optimisation, application expansion, and policy support" is relatively general. Suggestion refinement: Policy level: Which countries' subsidy models can be used for reference

Response 6: Thank you for pointing this out. We very much agree with your suggestion to refine the policy level in the text, specifically where to revise it - page 26, paragraph 1, line 1011, yellow highlighted.

 

Each reviewer's review comments were carefully responded to, and the review comments were always replied to in the paper.

Author Response File: Author Response.docx

Reviewer 2 Report

Comments and Suggestions for Authors

It is an interesting issue to review, but I would recommend including a scheme or making a specific reference to the classification related to the EMS techniques (typically rule-based, optimization, and learning methods).

I find a lack of order in the papers included in each subchapter. The classification (first rule-based, then optimization, and finally learning methods) could be one. Additionally, I find a lack of criteria or explanation for why those papers have been referenced (there are plenty of others in the literature) and not others. A representative publication for each EMS methodology (rule-based, optimization, and learning) could be one, but you could propose any other.

Specific comments:

• (Line 23): The advantages and disadvantages between direct and indirect types are explained in the introduction (around line 70). I consider it interesting to do something similar related to the three types of vehicles (passenger cars, buses, and heavy trucks) and also the three topologies according to power sources (fuel cell with battery, with UC, or with both of them). Some explanation related to the three topologies is given at the beginning of each chapter (2, 3, and 4), but a brief previous comparison in the introduction would be appreciated.

• Additionally, passenger car, bus, or heavy truck applications each have their own demands and limitations (different vehicle weight, torque and speed range, jerk limitations, autonomy requirement...). It would be appreciated to explain this a little bit to keep in mind the differences in each application.

• (Line 919): I agree that DRL is a promising algorithm for this purpose, but this should be somehow identified by the papers proposed in previous chapters. I do not see it so clearly. Additionally, other factors such as critical component degradation or efficiency maximization are usually considered rather than just SOC.

• (Lines 921 and 923): I totally agree that digital twin and FC heat waste use are really promising. Maybe some references or more mentions in the previous chapters should be included instead of just directly in the summary.

Author Response

Thank you very much for taking the time to review this manuscript. Please find the detailed responses below. The manuscript was further revised in terms of language and optimized for image typography, proportions, and so on.

 

Comments 1: I find a lack of order in the papers included in each subchapter. The classification (first rule-based, then optimization, and finally learning methods) could be one. Additionally, I find a lack of criteria or explanation for why those papers have been referenced (there are plenty of others in the literature) and not others. A representative publication for each EMS methodology (rule-based, optimization, and learning) could be one, but you could propose any other.

Response 1: Thank you for your valuable input. While many papers have been published related to energy management strategies by categorizing them, the manuscript has been completed using different fuel cell powertrains as a framework and considering different vehicle types. Because of your reminder, the manuscript does lack a description of the classification of the strategies, and the exact location of the revision - page 2, paragraph 1, line 63, yellow highlighting.

Also, regarding the research on energy management strategies, the relevant papers cited in the manuscript are almost recent research. The selection of relatively new research provides some reference to future energy management control strategies.

 

Comments 2: (Line 23): The advantages and disadvantages between direct and indirect types are explained in the introduction (around line 70). I consider it interesting to do something similar related to the three types of vehicles (passenger cars, buses, and heavy trucks) and the three topologies according to power sources (fuel cell with battery, with UC, or with both of them). Some explanation related to the three topologies is given at the beginning of each chapter (2, 3, and 4), but a brief previous comparison in the introduction would be appreciated.

Additionally, passenger car, bus, or heavy truck applications each have their own demands and limitations (different vehicle weight, torque and speed range, jerk limitations, autonomy requirement...). It would be appreciated to explain this a little bit to keep in mind the differences in each application.

Response 2: Thank you for pointing this out. We agree with your suggestion. We have revised the manuscript accordingly. A short comparison of the three topologies has been added to Chapter 1, specific revision location - page 2, paragraph 2, line 82, yellow highlighting.

In addition, there is a lack of differentiation notes regarding the different types of vehicles, which is the reason why the manuscript is discussed by different vehicle types. Specific revision location - page 4, paragraph 2, line 148, yellow highlight.

 

Comments 3: (Line 919): I agree that DRL is a promising algorithm for this purpose, but this should be somehow identified by the papers proposed in previous chapters. I do not see it so clearly. Additionally, other factors such as critical component degradation or efficiency maximization are usually considered rather than just SOC.

Response 3: Thank you for your keen insight into the importance of Deep Reinforcement Learning (DRL) in fuel cell hybrid systems. Your suggestion is very constructive and this paper clearly states in Section 5.2 that “Deep Reinforcement Learning (DRL) algorithm to dynamically adjust power allocation according to road conditions and state of charge (SOC) to adapt to complex working conditions”. This statement positions DRL as one of the core breakthrough directions for future intelligent energy management, which is highly consistent with your forward-looking view.

The feasibility of DRL algorithms has been verified through specific cases (e.g., optimization of fuel cell-battery power allocation by DDPG algorithm in Ref. [27], application of TD3 algorithm in health-aware management in Ref. [39]), which lays a practical foundation for the further application of DRL. Given that the current research focuses on the engineering validation of existing algorithms, the systematic application of DRL as a more cutting-edge optimization tool can be an important extension of future research. We will further explore the optimization strategy of DRL fused with physical models in our subsequent work to promote the intelligent development of fuel cell hybrid power systems.

 

Comments 4: (Lines 921 and 923):I totally agree that digital twin and FC heat waste use are really promising. Maybe some references or more mentions in the previous chapters should be included instead of just directly in the summary.

Response 4: Thank you for your valuable input. We fully agree with the prospective value of digital twin technology and fuel cell (FC) waste heat utilization in system optimization. In conjunction with the existing article structure, we note that:

(1) Section 5.2 of the paper has clearly stated “Combined with digital twin technology to realize dynamic monitoring of system life, and use of fuel cell waste heat to improve comprehensive energy efficiency.” (P25), which logically echoes the energy management strategies in the previous chapters. For example, the prediction of fuel cell degradation by dynamic planning in Section 2.3 Heavy Duty Truck Health Awareness Strategies (Ref. [43]) is essentially a preliminary application of the digital twin idea, and the model-based power decoupling approach in Section 4.1 Triple-Source Hybrid System Optimization (Ref. [73]) provides the technical basis for model construction of the digital twin.

(2) Waste heat recovery technology has been implicitly reflected in application scenarios in previous chapters: section 3.2 Thermal management of buses (Ref. [56]) mentions that “using the waste heat to warm the batteries may save 4% of energy”; section 4.2 The modeling of the degradation of hydrogen efficiency with the state of health (SOH) of the fuel cell in the three-source system for buses (Ref. [76]) indirectly reflects the impact of thermal management on the efficiency of the system.

Given that this paper focuses on the technology review of existing hybrid powertrains, the systematic study of digital twin and waste heat utilization as a frontier direction has been framed in the outlook section (Chapter 5). Subsequent studies can further explore the engineering applications of this technology in conjunction with the energy management strategies in the previous chapters. 

 

Each reviewer's review comments were carefully responded to, and the review comments were always replied to in the paper.

Author Response File: Author Response.docx

Reviewer 3 Report

Comments and Suggestions for Authors

Dear Authors,

 

 

General comment

I am a supporter of care for the nature and of action to protect environment. "Clean" electricity can significantly reduce our carbon footprint. The fuel cell is a good solution.

 

My selective comments are given as follows:

 

I have three doubts

1. Production

I am a supporter of innovative system. However, a sensible compromise should be reached while finding a solution. Currently, only "grey hydrogen" is widely available. "Clean hydrogen" is very expensive. So a compromise solution has to be chosen.

 

2. Quality

Hydrogen must have the right parameters. Contaminated hydrogen can destroy any installation.

 

 

3. Efficiency

The efficiency of a fuel cell depends on the power drawn. The higher the power had drawn the lower the efficiency of the cell. Vehicles in every city are exposed to this phenomenon. This is a serious limitation of this solution.

 

My Conclusion

1. I suggest discussing these three aspects.
2. Please carefully check all paper. Please revise the text; there are some errors to be fixed. I suggest eliminating these disadvantages.

Author Response

Thank you very much for taking the time to review this manuscript. Please find the detailed responses below. The manuscript was further revised in terms of language and optimized for image typography, proportions, and so on.

 

Comments 1: Production. I am a supporter of innovative system. However, a sensible compromise should be reached while finding a solution. Currently, only "grey hydrogen" is widely available. "Clean hydrogen" is very expensive. So a compromise solution has to be chosen.

Response 1: Thank you for your insight into the current state of hydrogen energy applications. The contradiction between “gray hydrogen dominance and clean hydrogen cost” that you mentioned is a key issue that needs to be balanced in the current development of the industry. Although this paper focuses on the technical optimization of fuel cell hybrid powertrain, the consideration of this contradiction is implicit in the strategy design - for example, one of the core objectives of the energy management strategies discussed in this paper (e.g., power allocation based on model predictive control, fuel cell health awareness algorithms, etc.) is to improve the efficiency of hydrogen utilization (e.g., reduction of equivalent hydrogen consumption, reduction of hydrogen cost, reduction of hydrogen consumption, etc.). （For example, one of the core objectives of the energy management strategies discussed in the paper (e.g., model-based predictive control of power allocation, fuel cell health state sensing algorithms, etc.) is to improve the efficiency of hydrogen energy utilization (e.g., lowering the equivalent hydrogen consumption, reducing the dynamic loss of fuel cells), to achieve the “efficiency-first” compromise under the existing hydrogen application scenarios, and to lay the technological foundation for the future scaled-up clean hydrogen application. In fact, as mentioned in the outlook section of the paper, the system optimization (e.g., thermal management synergy, multi-energy synergistic control) emphasized in the study not only meets the current cost constraints of grey hydrogen, but also prepares technical interfaces (e.g., more efficient electricity-hydrogen conversion, whole life cycle carbon emission optimization) for the coming of the clean hydrogen era. emissions optimization). We fully agree with your proposal for a “compromise solution”, which fits well with the “technically feasible - future-proof” framework of this paper, and thank you again for your professional advice.

 

Comments 2: Quality. Hydrogen must have the right parameters. Contaminated hydrogen can destroy any installation.

Response 2: Thank you for your valuable comments. The “hydrogen parameter control and pollution protection” you mentioned is indeed a key prerequisite for the safe operation of fuel cell systems, and this issue is of great significance in the research of fuel cell field. This paper focuses on the topology design of fuel cell hybrid power system, energy management strategy and the optimization of adaptability of different vehicle models. Although the hydrogen purification technology has not been developed in detail in the main text, the basic consideration of hydrogen quality has been implied in the underlying logic of the system design - for example, the optimization strategy of fuel cell durability discussed in the paper (e.g., limitation of For example, the fuel cell durability optimization strategies discussed in the paper (e.g., limiting power fluctuations, health monitoring, etc.) are essentially subsequent system optimization based on the premise of “assuming that the hydrogen parameters meet the operating standards”, while the front-end aspects such as hydrogen purification, storage, and impurity control belong to the key technologies complementary to the design of the power system, and the related contents have already been systematically developed in other studies in the field (e.g., hydrogen Impurity content and fuel cell life quantitative relationship).

 

Comments 3: Efficiency. The efficiency of a fuel cell depends on the power drawn. The higher the power had drawn the lower the efficiency of the cell. Vehicles in every city are exposed to this phenomenon. This is a serious limitation of this solution.

Response 3: Thank you for pointing this out. This phenomenon is indeed a key challenge that needs to be urgently addressed when fuel cells are applied to urban vehicles, and this issue is of great significance in guiding the design of systems under dynamic operating conditions. Although this challenge exists objectively, the continuous progress of related technologies is gradually breaking through this limitation. We fully agree with the problem you pointed out, and this direction provides an important optimization target for the subsequent technology development, thank you again for your valuable comments.

 

Comments 4: Please carefully check all paper. Please revise the text; there are some errors to be fixed. I suggest eliminating these disadvantages.

Response 4: Thank you for pointing this out. To further enhance the readability and rigor of the article, we have systematically optimized the language presentation, unified the terminology and sorted out the logic, etc. In addition, we have optimized the graphs and charts in the whole article. Thank you again for your guidance, these changes will help the article to better present the research results.

 

Each reviewer's review comments were carefully responded to, and the review comments were always replied to in the paper.

Author Response File: Author Response.docx