Communication Delay Prediction of DPFC Based on SAR-ARIMA-LSTM Model

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors 1. Many figures contain font sizes that are too small to read clearly. Figure 6 requires layout improvements, and Figure 9 lacks captions. 2. In Section 3.3, the authors compare two ARIMA-LSTM hybrid methods (Method 1 and Method 2) using RMSE/MAE. While Method 1 suggests better performance with higher LSTM weights, the paper fails to provide results for standalone ARIMA or LSTM models. Without these baselines, the claimed superiority of the hybrid approach remains unsubstantiated. 3. The study lacks comparisons with advanced models like Transformer-based architectures or Spatio-Temporal Graph Neural Networks (ST-GNNs), which are known for handling temporal and spatial dependencies. Why were these models not considered? 4. A table showing performance gains (ARIMA → ARIMA-LSTM → SAR-ARIMA-LSTM) would clarify contributions. 5. The paper omits critical preprocessing steps. How were missing values handled? Was interpolation or filtering applied? 6. Regarding the delay-sensitive nature of DPFC control, could the authors please comment on whether the 5-minute sampling resolution used in this study is sufficient to capture all relevant delay characteristics? 7. The distance-based threshold method (Eq. 2) lacks clarity on: How the threshold θ was selected, and why alternative weight definitions (e.g., communication load-based or device importance-weighted) were not explored. 8. The integration mechanism of spatial features extracted by SAR into the ARIMA-LSTM framework requires clearer explanation to enhance methodological transparency. 9. While the study effectively employs pseudo R² to assess SAR model configurations, the prediction performance evaluation relies solely on RMSE and MAE metrics. Incorporating additional standard metrics like R² (for variance explanation) and MAPE (for relative error assessment) would provide a more comprehensive and interpretable evaluation of model performance across different aspects. Recently, a paper proposed a Weighted Quality Evaluation (WQE)method combining these four metrics, which may be worth considering. If you find it interesting, please refer to “Zhou, Y., He, X., Montillet, J. P., Wang, S., Hu, S., Sun, X., ... & Ma, X. (2025). An improved ICEEMDAN-MPA-GRU model for GNSS height time series prediction with weighted quality evaluation index. GPS Solutions, 29(3), 1-19.”

 

Author Response

请参阅附件

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

This paper proposes the Communication Delay Prediction of DPFC Based on SAR-ARIMA-LSTM Model. After reviewing, there are some comments.

1. Why is the Spatial Autoregressive Model used for the prediction of the communication delay of the DPFC?
2. The coordination of the ARIMA and LSTM is ambiguous. It would be good to introduce mode clearly.
3. What are the residuals of the Communication Delay Prediction of DPFC? Could the LSTM predict the residuals precisely?
4. The structure of the Distributed Power Flow Controller should be introduced more clearly. The requirement of the Communication Delay of the DPFC should be clarified.
5. How are the parameters of the ARIMA model calculated? The solution method for the parameter estimation of the ARIMA model should be provided.

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

The main problem that the paper is addressing is not clear:
- What a Distributed Power Flow Controller is should be better introduced.
- Why is there a need for an algorithm for communication delay prediction? Why can it not be directly measured?
- The state-of-the-art analysis is highly insufficient.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

The authors have response to all my comments and can be accepted with slight modify.

Figure 6 is still in low quality.

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

Thanks for the revision. I have no further comments.

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

The authors partially addressed my concerns. Overall, the quality of the paper has been improved.

Author Response

Please see the attachment

Author Response File: Author Response.pdf