Encoder–Decoder-Based Velocity Prediction Modelling for Passenger Vehicles Coupled with Driving Pattern Recognition

Round 1

Reviewer 1 Report

This paper proposes an Encoder-decoder based velocity prediction modelling coupled with driving pattern recognition. Firstly, the driving pattern recognition model is established by K-means clustering algorithm and validated on test data. Then, a DPR-ED model is designed with the introduction of the Encoder-decoder structure. The DPR-ED model enables the simultaneous input of multiple temporal features to improve the prediction accuracy and stability.

In Section 2 it would be interesting to show the effect of the use of all 6 different driving cycles.

In Section 4 Figure 12 should be further explained. The axis of Counts in Figure 13 should have the same minimum and maximum value for both 13(a) and 13(b) figures.

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Reviewer 2 Report

1- There are many abbreviations in the manuscript. Though many could be guessed, a proper explanation can significantly improve readability.

For example, in the abstract: "... this paper proposes an Encoder-decoder (ED) ..." <-- add the abbreviation here between brackets

The manuscript could also benefit from including a separate list of abbreviations.

2- Line 158, using maximum velocity (v_max) as a feature can be problematic if there is noise in the speed measurement signal (as is the case for many real-world applications). Could the authors perhaps comment on using v_max of a filtered speed signal? Or the v_max value for a 5-second moving average?

3- Fig. 10 (especially Fig. 10.a), even though the predicted speed follows along well with actual velocity... the predicted speed seems quite noisy.

Did the authors examine the equivalent corresponding acceleration (and whether the vehicle is actually capable of changing speed at that rate)?

Not sure about the best way to address this, but it seems like an issue that ought to be brought up in the discussion.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf