Comparison among Active Front, Front Independent, 4-Wheel and 4-Wheel Independent Steering Systems for Vehicle Stability Control

Round 1

Reviewer 1 Report

The work is technically sound and provides a valuable contribution, and the manuscript clearly describes methods and outcomes.

Some minor revisions are advised ad follows.

The scope of the work is appropriate for the Journal, given the impact on recent automotive industry (e.g. steer-by-wire technology) or robotics. In light of this, some (more) comments on the matter are suggested, e.g. possible applications or extensions of the findings.

The organization of the paper can be (optionally) improved for clarity, including data on case study in Section II (which describes materials and methods) and dedicating Section III to results.

While non-traditional and independent steering systems have been widely investigated in research, few works investigates the performance of these systems with respect to vehicle stability. If possible, please include some comments to compare original findings with those of previous works focused on stability.

Please motivate the choice the 2-DOF bicycle model.

Some minor language corrections are necessary throughout the text. (Also, usually equations are cited as "Equation (#)" or "Eq. (#)" at the beginning of a sentence and "(#)" otherwise.)

Author Response

[1] The scope of the work is appropriate for the Journal, given the impact on recent automotive industry (e.g. steer-by-wire technology) or robotics. In light of this, some (more) comments on the matter are suggested, e.g. possible applications or extensions of the findings.

==> I think the results of this paper can be extended into autonomous driving, as mentioned in Conclusion. To emphasize that, the reference [36] was added in the revised manuscript. 

[36] Hu, C.; Wang, R.; Yan, F.; Integral sliding mode-based composite nonlinear feedback control for path following of four-wheel independently actuated autonomous vehicles, IEEE Transactions on Transportation Electrification, 2016, 2, 2, 221-230.

[2] The organization of the paper can be (optionally) improved for clarity, including data on case study in Section II (which describes materials and methods) and dedicating Section III to results.

==> Following the comments of the reviewer, Section II was divided into II and III. The subsection 2.3 was renamed into Section III in the revised manuscript.

3. Determination of Steering Angles from Lateral Tire Forces

[3] While non-traditional and independent steering systems have been widely investigated in research, few works investigates the performance of these systems with respect to vehicle stability. If possible, please include some comments to compare original findings with those of previous works focused on stability.

==> In Introduction, the survey on the previous researches on 4WIS were newly added in the revised manuscript. Moreover, the new references, [4],[5],[8],[9],[13],[14],[15],[16], [18],[24], and [36], were added in the revised manuscript.

There have been several approaches that applied 4WIS to vehicle stability control [10-17]. In the previous study, the control yaw moment was computed with sliding mode control [10]. To generate the control yaw moment in real vehicles, the steering angles of 4WIS, combined with braking and traction, were determined by quadratic optimization with equality constraints [10]. This is common approach in the previous work [20-23]. On the other hand, the steering angles of 4WIS were directly determined by a controller or kinematics [11-17]. In the research, LQ optimal and fuzzy logic parameter adjuster were adopted to determine the steering angles of 4WIS, and feedforward and PID control were adopted for TVD [11]. Like [10], the steering and traction devices were used as an actuator in the research. In another research, the steering angles of 4WIS or RWS and the traction torques of hub motors were calculated with μ-synthesis technique [12,13]. These researches have adopted the steering and the braking/traction as actuators for yaw moment generation [10-13]. This is called an integrated chassis control. On the contrary, some researches have adopted only the steering actuators. In the work of [14], the rear steering or toe angles in 4WIS system were determined by feedback with front steering angle and yaw rate signals. This is common in vehicle stability control with 4WS [18]. In the other research, PID control was adopted to determine the steering angles for position and kinematic constraint controller [15]. In another research, the steering angles of 4WIS were determined by kinematic relationship without controller design procedure for path tracking [16,17].

As given in the above reference survey, there have been two categories in applying 4WIS to vehicle stability control. The first is to combine 4WIS with the braking/traction devices. The second is to use only 4WIS. Among them, there have been little approaches to compare the performance of 4WIS with the other devices such as ESC and TVD or AFS, ARS and 4WS. In the previous research, AFS, ARS and 4WS were compared to one another in view of vehicle stability control [21]. However, there have been little approaches to compare 4WIS with AFS, FWIS and 4WIS. So, these steering actuators are compared with one another in terms of vehicle stability control in this paper. 

==> In Section 4, new findings on 4WIS were added into the revised manuscript.

The magnitudes of the steering angle of 4WS and 4WIS are smaller than those of AFS and FWIS, as shown in Figure 5. This is caused by the use of RWS and RWIS in 4WS and 4WIS, respectively. Moreover, the direction of RWS is opposite to that of AFS in case of 4WS and 4WIS, as shown in Figure 5. This have been widely adopted method in applying RWS to vehicle stability control [14,18,32]. The results given in Figure 5 show the advantage of FWIS and 4WIS over AFS and 4WS, respectively. For FWIS and 4WIS, the steering angle of front inner wheel should be larger than that front outer one when cornering. FWIS and 4WIS can give better control performance than AFS and 4WS because it can generate larger lateral tire forces, as shown in Figure 5.

[4] Please motivate the choice the 2-DOF bicycle model.

==> The reason why 2-DOF bicycle model was used is given in the first paragraph of the subsection 2.1.

For controller design, a vehicle model is needed. Several types of vehicle models have been used for controller design. Typical vehicle models for vehicle stability control are 3-DOF planar and 2-DOF bicycle ones. 3-DOF planar model has longitudinal, lateral and yaw dynamics [24]. So, a tire model, which can calculate the longitudinal and lateral tire forces, is needed. On the other hand, 2-DOF bicycle model has no longitudinal dynamics. Only a lateral tire force model is needed. Moreover, it is much simpler than 3-DOF planar model [20-23]. For the reason, the most frequently used model in vehicle stability control is a 2-DOF bicycle model. So, a 2-DOF bicycle model is used to design a yaw moment controller in this paper.  

[5] Some minor language corrections are necessary throughout the text. (Also, usually equations are cited as "Equation (#)" or "Eq. (#)" at the beginning of a sentence and "(#)" otherwise.)

==> The author followed the instruction as given in the manuscript template of Electronics Journal. So, I have unified the notation “Equation (#)“ regardless of its position.

Reviewer 2 Report

In the manuscript, a vehicle stability controller is designed. In terms of vehicle stability control, AFS, FWIS, 4WS and 4WIS are compared with each other. And the result shows that independent steering systems can improve the vehicle stability control. Although this research has certain practical significance, I don't think it is innovative enough. Meanwhile, there are also other imperfections in the work. I hope that my comments would be useful for improving the quality of the paper. Some of the detailed comments are as follows:

1. The recent references cited in the paper are not comprehensive enough and cannot reflect the author's professionalism in the field.
2. It is impossible to prove the feasibility and validity of the calculation method only by simulation experiment.
3. Language expression still needs to be improved, such as line79，line147 and so on.

Author Response

[1] The recent references cited in the paper are not comprehensive enough and cannot reflect the author's professionalism in the field.

==> In Introduction, the survey on the previous researches on 4WIS were newly added in the revised manuscript. Moreover, the new references, [4],[5],[8],[9],[13],[14],[15],[16], [18],[24], and [36], were added in the revised manuscript.

There have been several approaches that applied 4WIS to vehicle stability control [10-17]. In the previous study, the control yaw moment was computed with sliding mode control [10]. To generate the control yaw moment in real vehicles, the steering angles of 4WIS, combined with braking and traction, were determined by quadratic optimization with equality constraints [10]. This is common approach in the previous work [20-23]. On the other hand, the steering angles of 4WIS were directly determined by a controller or kinematics [11-17]. In the research, LQ optimal and fuzzy logic parameter adjuster were adopted to determine the steering angles of 4WIS, and feedforward and PID control were adopted for TVD [11]. Like [10], the steering and traction devices were used as an actuator in the research. In another research, the steering angles of 4WIS or RWS and the traction torques of hub motors were calculated with μ-synthesis technique [12,13]. These researches have adopted the steering and the braking/traction as actuators for yaw moment generation [10-13]. This is called an integrated chassis control. On the contrary, some researches have adopted only the steering actuators. In the work of [14], the rear steering or toe angles in 4WIS system were determined by feedback with front steering angle and yaw rate signals. This is common in vehicle stability control with 4WS [18]. In the other research, PID control was adopted to determine the steering angles for position and kinematic constraint controller [15]. In another research, the steering angles of 4WIS were determined by kinematic relationship without controller design procedure for path tracking [16,17].

As given in the above reference survey, there have been two categories in applying 4WIS to vehicle stability control. The first is to combine 4WIS with the braking/traction devices. The second is to use only 4WIS. Among them, there have been little approaches to compare the performance of 4WIS with the other devices such as ESC and TVD or AFS, ARS and 4WS. In the previous research, AFS, ARS and 4WS were compared to one another in view of vehicle stability control [21]. However, there have been little approaches to compare 4WIS with AFS, FWIS and 4WIS. So, these steering actuators are compared with one another in terms of vehicle stability control in this paper. 

[2] It is impossible to prove the feasibility and validity of the calculation method only by simulation experiment.

==> I agree the reviewer’s comment. However, the purpose of this paper is to compare the performance of AFS, FWIS, 4WS and 4WIS in terms of maneuverability and lateral stability. I think this can be validated through simulation on a vehicle simulation package without experiments.

[3] Language expression still needs to be improved, such as line79，line147 and so on.

==> The grammatical errors in line79 and line147 were corrected.

==> The author has checked the grammatical errors in the revised manuscript.