A Dynamic Inertia Control Method for a New Energy Station Based on a DC-Driven Synchronous Generator and Photovoltaic Power Station Coordination
Round 1
Reviewer 1 Report
Comments and Suggestions for Authors1) Lines 59, 62, 66, 70, 76, etc.
The authors use the phrase 'Literature [N]' many times. What do you mean? This sentence is not English.
2) The reference to [1] is missing from the manuscript.
3) Line 120: The notation in the last paragraph of the first section contradicts the rubrication of the article. The article does not contain parts numbered with Latin numerals.
4) The general rule for formatting English text with formulas is the following.
A formula is a normal part of a sentence and all the rules of punctuation in English should apply to such sentences.
In particular, commas should be after formulas (1), (2), (4), (6), (7) and other similar places before `where`, which should be with a lowercase letter.
Starting a sentence with the word `Where` is bad style.
A sentence cannot begin with a formula.
Two formulas in a row should be separated by commas. These are formulas (6) and (7); (9) (double numbering!); (15) and (16), etc.
5) The periods should be after formulas (3), (5), (8), (9) and other similar places.
6) Line 132: Comma should be after `inertia of the rotor`. This is an enumeration. See other similar places too.
7) Lines 135, 137: The mathematical symbols D and D are different symbols. Similar situation with the symbols t and t in formula (3) and with the symbols H and H below.
Line 135 and Line 143: Differential d and d must be written the same way. Moreover, d/dt means total (Euler) derivative and the reader should not be confused.
Pairs of indices `m` and `m`, `r` and `r`, `i` and `i` must be the same. See other similar situations as well.
Indices must be written below the main level of the formula! (Lines 132, 138, 144, 146, etc.)
See other similar places in the text.
8) Line 137: The formula needs to be corrected: ∆ωm .
9) Lines 144, 145: What does the phrase `Ta is acceleration and deceleration torque` mean?
10) What is the electrical torque?
11) What is Sr , Pm, Pe in (5) ?
12) Line 174: What does `0` mean? Is it the initial moment of time or an index? How then to understand formula (10)?
13) The structure of Section 2 is unacceptable. It is necessary to formulate the general statement of the problem. And then describe the adopted model. You start with the formula for the kinetic energy of one solid. The reader does not understand what you are talking about. Why are you writing about the kinetic energy of rotation if this value is not considered further? A radical restructuring of the article is required.
14) It is necessary to formulate the Cauchy problem or boundary value problem for (8).
15)The exp argument in formula (9) needs to be corrected.
16) Formula (11) needs justification.
17) Formula (12) needs justification. It is unclear - is this an equation or an identity?
18) What does `x` in (14) mean? Is it a vector product?
19) What is `p` in (15) ?
20) How does (17) follow from Figure 3? What modeling tool is used?
21) How are the right parts in (17) and (19) related?
22) Formulas (23) and (24) contain symbols unknown to the reader.
23) The way the material is presented in this section needs to be completely revised. It is unclear what the authors want to convey to the reader with formulas (29), (30).
24) Where are the results of training and testing the neural network?
25) The frequency f first appears in Figure 9. This requires clarification.
Comments on the Quality of English Language
Numerous examples are given in the main review.
Author Response
1) Lines 59, 62, 66, 70, 76, etc.
The authors use the phrase 'Literature [N]' many times. What do you mean? This sentence is not English.
Response: Thanks for your suggestion. The article changed ' Literature [ N ] ' to ' Reference [ N ] ',and the modification has been marked in red.
2) The reference to [1] is missing from the manuscript.
Response: Thanks for your suggestion. The article supplements the reference [ 1 ], and the modification has been marked in red.
3) Line 120: The notation in the last paragraph of the first section contradicts the rubrication of the article. The article does not contain parts numbered with Latin numerals.
Response: Thanks for your suggestion. The article changes the Roman numerals to Arabic numerals, which corresponds to the chapter of the article, and the modification has been marked in red.
4) The general rule for formatting English text with formulas is the following.
A formula is a normal part of a sentence and all the rules of punctuation in English should apply to such sentences.
In particular, commas should be after formulas (1), (2), (4), (6), (7) and other similar places before `where`, which should be with a lowercase letter.
Starting a sentence with the word `Where` is bad style.
A sentence cannot begin with a formula.
Two formulas in a row should be separated by commas. These are formulas (6) and (7); (9) (double numbering!); (15) and (16), etc.
Response: Thanks for your suggestion. The article changed ' Where ' to ' where '. Continuous formulas have been separated by commas. The modification has been marked in red.
5) The periods should be after formulas (3), (5), (8), (9) and other similar places.
Response: Thanks for your suggestion. The article corrects the punctuation. The modification has been marked in red.
6) Line 132: Comma should be after `inertia of the rotor`. This is an enumeration. See other similar places too.
Response: Thanks for your suggestion. This paper corrects the punctuation marks of the full text. The modification has been marked in red.
7) Lines 135, 137: The mathematical symbols D and D are different symbols. Similar situation with the symbols t and t in formula (3) and with the symbols H and H below.
Line 135 and Line 143: Differential d and d must be written the same way. Moreover, d/dt means total (Euler) derivative and the reader should not be confused.
Pairs of indices `m` and `m`, `r` and `r`, `i` and `i` must be the same. See other similar situations as well.
Indices must be written below the main level of the formula! (Lines 132, 138, 144, 146, etc.)
See other similar places in the text.
Response: Thanks for your suggestion. In this paper, the letters are unified in full text, and the subscripts ' m ', ' r ' and ' i ' are unified in format. The modification has been marked in red.
8) Line 137: The formula needs to be corrected: ∆ωm .
Response: Thanks for your suggestion. The formula is modified in this paper. The modification has been marked in red.
9) Lines 144, 145: What does the phrase `Ta is acceleration and deceleration torque` mean?
Response: Thanks for your suggestion. Ta is the acceleration torque. The modification has been marked in red.
10) What is the electrical torque?
Response: Thanks for your suggestion. Electrical torque is the rotational torque generated by the interaction between electric energy and magnetic field. The modification has been marked in red.
11) What is Sr , Pm, Pe in (5) ?
Response: Thanks for your suggestion. pm is mechanical power, pe is the electric power, Sr is rated power .The modification has been marked in red.
12) Line 174: What does `0` mean? Is it the initial moment of time or an index? How then to understand formula (10)?
Response: Thanks for your suggestion. The subscript ( 0 ) refers to the fundamental frequency of the system under rated conditions, and the subscript ( 1 ) indicates that the parameter is the calculation result after per-unit normalization based on the system reference capacity or reference conditions. Formula 10 can be further understood by Figure 4. The modification has been marked in red.
13) The structure of Section 2 is unacceptable. It is necessary to formulate the general statement of the problem. And then describe the adopted model. You start with the formula for the kinetic energy of one solid. The reader does not understand what you are talking about. Why are you writing about the kinetic energy of rotation if this value is not considered further? A radical restructuring of the article is required.
Response: Thanks for your suggestion. The paper first clarifies the general description of the problem, and adds a model of the influence of rotational kinetic energy on frequency control in the later chapters of the article. The modification has been marked in red.
14) It is necessary to formulate the Cauchy problem or boundary value problem for (8).
Response: Thanks for your suggestion. In this paper, the boundary value of the formula is set. The modification has been marked in red.
15)The exp argument in formula (9) needs to be corrected.
Response: Thanks for your suggestion. The exponential parameter of Formula 9 is modified. The modification has been marked in red.
16) Formula (11) needs justification.
Response: Thanks for your suggestion. If large-scale photovoltaic is integrated into the power system, this part of the photo-voltaic power station cluster is taken as a whole. At this time, the inertia Hsys(1) after grid connection can be expressed as : formula(14)
17) Formula (12) needs justification. It is unclear - is this an equation or an identity?
Response: Thanks for your suggestion. This paper proposes a control strategy that uses a shared energy storage device to compensate for the weakening of photovoltaic grid-connected inertia. The control topology is shown in Figure 1.
Fig. 1 Inertia compensation model of photovoltaic power station
Hwfj 、Swfj and Ej in the diagram represent the virtual inertia, rated capacity and kinetic energy weakening of the photovoltaic power station j ( j= 1,2...L ) ; ΔEESS、H ESS and SESS are the compensation kinetic energy, virtual inertia and rated capacity of shared energy storage, respectively.
The modification has been marked in red.
18) What does `x` in (14) mean? Is it a vector product?
Response: Thanks for your suggestion. X represents the product
19) What is `p` in (15) ?
Response: Thanks for your suggestion. There is a non-standard writing of the formula in the text. The formula is corrected to [18][19].
The modification has been marked in red.
20) How does (17) follow from Figure 3? What modeling tool is used?
Response: Thanks for your suggestion. From the system transfer function in the red box in Figure 3, Formula 17 can be obtained and modeled by MATLAB. The modification has been marked in red.
21) How are the right parts in (17) and (19) related?
Response: Thanks for your suggestion. The formula 17 is the system output frequency formula, and the formula 19 is the power instruction generated when the re-frequency changes. The system frequency is controlled by adjusting the system power. The modification has been marked in red.
22) Formulas (23) and (24) contain symbols unknown to the reader.
Response: Thanks for your suggestion. netj represents the input weighted sum of the jth neuron. x Represents the input vector. cj Represents the center vector of the jth hidden layer neuron. .bj represents the width parameter of the jth hidden layer neuron. The modification has been marked in red.
23) The way the material is presented in this section needs to be completely revised. It is unclear what the authors want to convey to the reader with formulas (29), (30).
Response: Thanks for your suggestion. The original expression of the article is wrong. Formulas ( 29 ) and ( 30 ) are obtained by discretization of ( 27 ) and ( 28 ). Because when the controller is used for control, the control algorithm needs to use discontinuous data for modeling. Therefore, it is expressed by formulas ( 29 ) and ( 30 ). The modification has been marked in red.
24) Where are the results of training and testing the neural network?
Response: Thanks for your suggestion. The article supplements the neural network training and test results.The number of model iterations is set to 200 times, and the network is initialized before training. The learning rate is 0.0001 and the batch is 256. Figure 11 shows how the accuracy of the training set and the test set changes with the number of iterations. The final accuracy of the training set and the verification set is 99.49 % and 99.09 %, respectively.
The modification has been marked in red.
25) The frequency f first appears in Figure 9. This requires clarification.
Response: Thanks for your suggestion. In the simulation process, the parameter
f=ω/2π. The modification has been marked in red.
Reviewer 2 Report
Comments and Suggestions for AuthorsThis paper critically examines the challenge of inadequate inertia provision in grid-integrated photovoltaic (PV) power plants. To bolster system stability and frequency regulation, the authors introduce an innovative dynamic inertia control strategy. This strategy combines a direct current (DC)-coupled synchronous generator with virtual synchronous control of PV inverters. The paper further develops a station-level coordinated control model that integrates physical and virtual inertia sources. This model incorporates neural network-based predictive control to optimise inertia parameters dynamically. Simulation results obtained using MATLAB/Simulink demonstrate the efficacy of the proposed methodology in enhancing frequency response and system resilience under disturbance conditions. Notable strengths of this paper include the comprehensive mathematical modelling of system dynamics and inertia mechanisms, as well as the integrated utilisation of physical inertia from DC-driven synchronous generators and virtual inertia emulated by PV inverters. Before acceptance, the authors must address the following:
- Although the authors cite many references, the review lacks critical analysis and clear linkage to how the proposed method improves over existing techniques in the literature review (Introduction) section.
- The organisation of the content, especially in the early sections (Introduction and Related Work), lacks coherence. Some sections present overlapping content or diverge without clear transitions.
Example: Several sections (e.g., Sections 3–5) mix modelling, control strategy, and simulation setup without a clear distinction between them. The purpose of each section is not well defined.
- Some equations are poorly formatted, and figure captions lack sufficient explanation. Figures such as control block diagrams and system configurations should be made clearer with labelled components (Figs 1, 2 and 5)
- Some variables and constants need to be described in the text as PD, PFFC, PSIC, KFFC, KSIC on page 4.
- Parameters used in the simulation (e.g., inertia constants, controller gains, neural network weights) are presented with little or no explanation of how they were selected or optimised.
- Some figures (e.g., frequency deviation graphs) are not properly labelled with axis units or legends.
- Figures 9 to 18 display performance metrics, but statistical analysis (like overshoot%, settling time, RMSE) is missing.
- The paper contains numerous grammatical errors and awkward sentence constructions, which impede clarity and reduce the overall readability. Thorough proofreading or professional editing is necessary.
Examples:
"Improve the robustness of the control system." (Incomplete sentence)
"Through theoretical analysis and simulation verification, under the constraint of frequency response as the main dynamic index." ( lacks a complete thought).
Comments on the Quality of English LanguageThe paper contains numerous grammatical errors and awkward sentence constructions, which impede clarity and reduce the overall readability. Thorough proofreading or professional editing is necessary.
Author Response
his paper critically examines the challenge of inadequate inertia provision in grid-integrated photovoltaic (PV) power plants. To bolster system stability and frequency regulation, the authors introduce an innovative dynamic inertia control strategy. This strategy combines a direct current (DC)-coupled synchronous generator with virtual synchronous control of PV inverters. The paper further develops a station-level coordinated control model that integrates physical and virtual inertia sources. This model incorporates neural network-based predictive control to optimise inertia parameters dynamically. Simulation results obtained using MATLAB/Simulink demonstrate the efficacy of the proposed methodology in enhancing frequency response and system resilience under disturbance conditions. Notable strengths of this paper include the comprehensive mathematical modelling of system dynamics and inertia mechanisms, as well as the integrated utilisation of physical inertia from DC-driven synchronous generators and virtual inertia emulated by PV inverters. Before acceptance, the authors must address the following:
- Although the authors cite many references, the review lacks critical analysis and clear linkage to how the proposed method improves over existing techniques in the literature review (Introduction) section.
Response: Thanks for your suggestion. The article further modifies the introduction, adds the corresponding reference comparison, gives a clear explanation of the shortcomings of the existing literature research, and highlights the innovation of this article.
The above research has greatly improved the stability of the system and has a good reference value. But most of them take energy storage as the main body. The frequency signal under grid disturbance is used as input to improve the frequency modulation strategy. The inertia weakening caused by new energy grid connection is not quantified. There is no effective compensation for inertia weakening. So that the inertia weakening problem still exists.
The modification has been marked in red.
- The organisation of the content, especially in the early sections (Introduction and Related Work), lacks coherence. Some sections present overlapping content or diverge without clear transitions.
Example: Several sections (e.g., Sections 3–5) mix modelling, control strategy, and simulation setup without a clear distinction between them. The purpose of each section is not well defined.
Response: Thanks for your suggestion. Thanks to the expert opinion, the article re-sorts the coherence of the content, and the verification of the corresponding control strategy is added to the simulation results. Each section is redefined and the corresponding explanation is added. The modification has been marked in red.
- Some equations are poorly formatted, and figure captions lack sufficient explanation. Figures such as control block diagrams and system configurations should be made clearer with labelled components (Figs 1, 2 and 5)
Response: Thanks for your suggestion. In this paper, the formula is standardized and the graphics are standardized. The modification has been marked in red.
- Some variables and constants need to be described in the text as PD, PFFC, PSIC, KFFC, KSIC on page 4.
Response: Thanks for your suggestion.
PD represents the reference power value on the load side .PFFC : The power value repre-senting the frequency change control. .PSIC : Indicates the power value of inertial control.KD is the damping factor, its physical meaning represents the electric load that changes its active power by frequency change. KFCC=KFCC(t-t0) is the proportional control coefficient of frequency control. KSIC=KSIC(t-t0) is the proportional control coefficient of inertial control.
The modification has been marked in red.
- Parameters used in the simulation (e.g., inertia constants, controller gains, neural network weights) are presented with little or no explanation of how they were selected or optimised.
Response: Thanks for your suggestion.
The corresponding simulation parameters are added in this paper.
system |
parameter |
taking values |
photovoltaic |
apparent power S |
500MVA |
Capacitance Cdc |
3.4 mF |
|
Inertia J |
3.6×104kg m2 |
|
direct current voltage Udc |
1000V |
|
damping coefficient Kd |
1.2×104 |
|
control parameter Kc |
800 |
|
proportional gain Kp |
0.5 |
|
Integral gain Ki |
100 |
|
electrified wire netting |
output power P |
320MW |
Parameters of inertia 1/D |
1.3×10-5 |
|
Inductance L |
5mH |
|
Capacitance C |
50uF |
|
Voltage Uabc |
1100V |
The modification has been marked in red.
- Some figures (e.g., frequency deviation graphs) are not properly labelled with axis units or legends.
Response: Thanks for your suggestion. In this paper, the simulation diagram is normalized. The modification has been marked in red.
- Figures 9 to 18 display performance metrics, but statistical analysis (like overshoot%, settling time, RMSE) is missing.
Response: Thanks for your suggestion. The article further increases the analysis of simulation results.
In the case of sudden increase and sudden decrease of load power. When the DC fre-quency droop control and inertia control are not put into operation, the DC system does not respond to the system frequency change, the DC transmission power remains un-changed, and it cannot participate in the system frequency modulation. At this time, the frequency characteristic of the sending end system is the worst. When only inertia control is used, the maximum deviation value ( the absolute value of the difference between the minimum value or the maximum value and the rated value ) of the frequency of the send-ing end system is improved. However, the steady-state value of the system frequency is not improved. When frequency coordinated control is adopted, the maximum deviation value and steady-state value of the frequency of the sending end system are improved the most. 0.39 Hz and 0.10 Hz respectively when the load increases suddenly. When the load is suddenly reduced, it is 0.41 Hz and 0.09 Hz respectively. And the system tends to be stable after 35 s.
The modification has been marked in red.
- The paper contains numerous grammatical errors and awkward sentence constructions, which impede clarity and reduce the overall readability. Thorough proofreading or professional editing is necessary.
Examples:
"Improve the robustness of the control system." (Incomplete sentence)
"Through theoretical analysis and simulation verification, under the constraint of frequency response as the main dynamic index." ( lacks a complete thought).
Response: Thanks for your suggestion. The sentences of the article are further standardized. The modification has been marked in red.
Reviewer 3 Report
Comments and Suggestions for AuthorsThe authors propose a photovoltaic power station controlled by synchronous generator and virtual synchronous power generation. They establish a station-level dynamic inertia control model with synchronous machine and inverter control parameters coordinated, while setup a cooperative inertia control method of photovoltaic grid-connected inverter and synchronous machine. Then, they carry out the influence of inertia on the system frequency, and frequency optimization of grid-connected parameter optimization of photovoltaic station based on inertia control. Aiming at the grid-connected control parameters, the inertia control parameter setting method of photovoltaic station is carried out. The paper is meanful for the grid-connection network. The paper should compare with brainIoT: brain-like productive services provisioning with federated learning in industrial IoT, reliable low-latency routing for VLEO satellite optical network: a multiagent reinforcement learning approach, HAP networking enables highly reliable space-air-ground optical interconnect: an integrated network perspective. The simulations should give in more details.
Author Response
The authors propose a photovoltaic power station controlled by synchronous generator and virtual synchronous power generation. They establish a station-level dynamic inertia control model with synchronous machine and inverter control parameters coordinated, while setup a cooperative inertia control method of photovoltaic grid-connected inverter and synchronous machine. Then, they carry out the influence of inertia on the system frequency, and frequency optimization of grid-connected parameter optimization of photovoltaic station based on inertia control. Aiming at the grid-connected control parameters, the inertia control parameter setting method of photovoltaic station is carried out. The paper is meanful for the grid-connection network. The paper should compare with brainIoT: brain-like productive services provisioning with federated learning in industrial IoT, reliable low-latency routing for VLEO satellite optical network: a multiagent reinforcement learning approach, HAP networking enables highly reliable space-air-ground optical interconnect: an integrated network perspective. The simulations should give in more details.
Response: Thanks for your suggestion.
The article compares the algorithm with the corresponding article. The article adds the corresponding simulation results. The modification has been marked in red.
Round 2
Reviewer 2 Report
Comments and Suggestions for AuthorsThe authors have done everything I requested in my comments. I have no further comments.