Determining the Proper Times and Sufficient Actions for the Collision Avoidance of Navigator-Centered Ships in the Open Sea Using Artificial Neural Networks

Round 1

Reviewer 1 Report

1. The paragraph starting from line 177 is used to illustrate the situation of 2-ship crossing encounter, and is drawn in Figure 1. This article explains that there will be some azimuth and distance parameters, but the parameters cannot correspond to the features in Figure 1. For example, Veering point and reach point should be defined in words and marked on the figure.

2. In line 185, how to define final modified collision avoidance action?

3. The definition of DPCA and TPCA is not clearly stated, and it is easy to be confused with TCPA or DCPA.

4. In line 184, Ros is set to 1.5, which is the median of the reach point range, how to get it?

5. Equation 6 and 7 are not marked, xto and yto of Equation 7 are defined, thereare words in Korean, in line 284

6. The relative orientation ∅ and undefined in Line 245, is it Φ in Equation 6.?

7. The equation below line 237; m and D should be drawn on Figure 2, which can be more clearly explained

8. line 293, Gauss–Nowton?

9. Please explain why 1000 epochs are used.

10. The model error may cause the control program to be different, error 2.34 miles, whether the collision avoidance strategy has changed (such as high-angle crossing, or small-angle crossing, or overtaking state). Please describe the possible impact.

11. The titles of vertical axis and horizontal axis of Figure 6 to Figure 11 should be clearly marked

12. The data in Table 3 is generated by the formula in Section 2.3. Why do we need to practice the ANN model? D and alpha can be obtained directly from the formula

13. The writing style of the whole article is not easy to read. You should check carefully, grammar, semantics, vocabulary, etc., or hire a native English speaker to revise it. It is recommended that re-evaluation after revision

The writing style of the whole article is not easy to read. You should check carefully, grammar, semantics, vocabulary, etc., or hire a native English speaker to revise it. It is recommended that re-evaluation after revision

Author Response

Thank you for your good comments. Based on your comments, we have revised the entire manuscript by rechecking the grammar, meaning, semantics, vocabulary by proofreading service of native English editing company. please see attachments.

Author Response File: Author Response.pdf

Reviewer 2 Report

First of all, the paper needs to be improved; there are a lot of editorial errors.
The topic indicates the research is made in Open Sea, while the verification was made in simulations software.


Bayesian networks are a widely-used class of probabilistic graphical models. They consist of two parts: structure and parameters. The structure is a directed acyclic graph (DAG) that expresses conditional independencies and dependencies among random variables associated with nodes. The parameters consist of conditional probability distributions associated with each node. A Bayesian network is a compact, flexible, interpretable representation of a joint probability distribution. It is also an useful tool in knowledge discovery as directed acyclic graphs allow the representation causal relations between variables. Typically, a Bayesian network is learned from data.

The question is what Authors would like to present in Fig. 3 Structure of a Bayesian Artificial Neural Network (BRANN)?

In general, paper needs major improvements.

Many abbreviations (CPA – in line 24, IMO in line 109, etc. ) are not defined before there are used the first time.
In line 248, the deflection angle is written in an inappropriate format.
The citations are made in a different ways; please unify them.
The reference needs to be improved according to the journal.
Not all equations are numbered.

Author Response

Thank you for your good comments. Based on your comments, we have revised the entire manuscript by rechecking the grammar, meaning, semantics, vocabulary by proofreading service of native English editing company. please see attachments

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

The authors respond to all proposed suggestions or improvements.

Moderate editing of English language required

Reviewer 2 Report

The article has been corrected in many parts and some important issues have been clarified. The article in its current form can be published.