Voltage Control for Active Distribution Networks Considering Coordination of EV Charging Stations

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The paper presents a timely and innovative solution for voltage control in EV-integrated ADNs. The fusion of data-driven RC estimation, uncertainty-aware optimization, and decentralized allocation holds significant promise. However, methodological ambiguities and lack of comparative benchmarks weaken the current presentation. Major revision is required by addressing these concerns.

1. The absence of comparative analysis with state-of-the-art methods, e.g., [15], [17], [20] makes claimed advantages over single-interval RC evaluation and centralized allocation unsubstantiated.
2. The RVM uncertainty modeling in Section 3.2.5 lacks critical implementation details: Specific input features for "holiday effects" and measurement methodologies for "charging fulfillment proportions" require clarification.
3. The SOCP relaxation gap for DistFlow equations remains unquantified, casting doubt on solution optimality despite constraint (32).
4. Practical implementation challenges are overlooked, particularly communication latency impacts on broadcast control convergence and ESS degradation effects omitted from the cost function, Eq. 36.
5. erminology inconsistency persists, "RC" vs. "adjustable power range", undefined "VOS" acronym at first use in Section 3.2.2.
6. Hyperparameter selection rationale is inadequate e.g., λ=0.01 in Eq. 11, α in Eq. 49, requiring ablation studies or sensitivity analysis.
7. Figure 14 would benefit from violation statistics such as duration/magnitude reduction to quantitatively demonstrate improvement.
8. The EV behavior model reliance on [29]'s 2017 data warrants justification given evolving EV usage patterns.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

Active Distribution Networks (ADNs) are increasingly affected by the widespread adoption of PV generation and EVs. Their variability may lead to voltage violations that compromise the safe operating area of ADNs. In this article, the authors propose leveraging the flexibility provided by EV Charging Stations (EVCS) through: (1) a state-driven margin algorithm and an MLP model to predict the response capability (RC) of EVCSs; (2) an RVM-based model to establish confidence intervals for EV availability; and (3) an SOCP-based voltage regulation framework combined with (4) a broadcast control scheme that distributes dispatch signals from the EVCS to EVs and ESS.

The topic of the article is relevant and of clear contemporary interest. The paper is clear, well-structured, and technically sound. The methodology is reproducible, the experiments are appropriate to test the hypotheses, and the results are well presented. However, the article contains some minor errors that should be addressed prior to publication:

1. Some acronyms are defined in the abstract but their definition in the body of the article are missing.
2. In line 121, you might want to consider whether it would be more accurate to write (i,j)∈L instead of (i,j)⊆L, since (i,j) represents an element rather than a subset.
3. In Figure 3, the subscript “i” of the dis/charging margin has not been defined. Please unify the notation. Moreover, Δt appears in this figure but is not defined until line 281; it may be clearer to introduce its meaning earlier.
4. In line 190 you might actually be referring to the segment “BC” instead of “CF.”
5. t_out​ is not defined in equation (2).
6. In line 212, the states “state(t) = 1,2,3 […] charging, idle, and discharging, respectively” do not correspond to those in equations (5) and (6), where 1: charging, 2: discharging, 3: idle.
7. In equations (5) and (6), the variable t could perhaps be replaced by t_c​. Otherwise, please define t.
8. In equation (9), you use u_i​, while in the following line you refer to μ_i​.
9. The analytical derivation of the dis/charging margins from equations (7) and (5–6) does not seem to be computationally demanding. It would be helpful if the authors could further clarify the advantages of employing the MLP-based model compared to the analytical calculation.
10. In equation (39), P_i,c^max⁡ seems to refer to P_c,m, defined in line 162, at bus i. Unifying the notation would improve the clarity of the article.
11. In equation (42), P_d^e​ and P_c^e​ have not been defined.
12. In Table 1, the units should be separated from the values.
13. In Figures 5 to 15, the text is hardly legible, the image resolution could be improved, and the aspect ratio of some of them have been modified. Please consider using vector graphics.
14. After the title of each reference, “[J]” or “[C]” appears written.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

I have no more comments.