Study of Pinion Gear Support Structure of Electric Power Steering System

Takashi Miyoshi (Honda R&D)

This study investigated the mechanical performance required for the pinion gear support structure, including the torsion bar, in an EPS (Electric Power Steering) system to achieve a direct steering feel. The results clearly showed that the linearity of the side force and displacement acting on the pinion gear contributes to the direct steering feel. Therefore, we propose a new structure that can contribute to enhancing steering feel.

Effects of Equivalent Steering Torque Characteristics Derived from the Coupling Structure of Steering System and Vehicle Dynamic Characteristics on Handling Quality Evaluation

Jun Ishio (Honda R&D)･Masato Abe･Ikuo Kushiro･Masaki Yamamoto･Reon Suzuki･Makoto Yamakado (Kanagawa Institute of Technology)

A coupled formulation of the steering system with assist control and vehicle dynamic characteristics is presented, and it is shown that the steering system can be reduced from two to one degree of freedom via internal forces, even in the presence of nonlinear EPS assist control. The effects of the derived equivalent steering torque characteristics on driver handling quality evaluation are confirmed with driving simulator.

Proposal of Parametric Road Feel Reproduction Based on Phase Coupling in Tire-Order Vibration Sidebands

Takuya Suzuki (Honda R&D)

To reproduce road feel in steer-by-wire vehicles, phase difference analysis was applied to steering vibration during driving. The sidebands of tire rotational order vibration were found to be phase-coupled to the carrier at 180°, indicating they represent amplitude modulation in the FFT domain. A parametric excitation model preserving this coupling structure was proposed, demonstrating improved signal reproducibility compared to conventional methods.

Development of Steering Control for Steer-by-Wire Systems Using a Hierarchical Gradient Method
-(Part 1)-

Rei Fukai･Takashi Furuichi･Atsushi Tsubouchi (Honda R&D)･Hideaki Shibue･Hideyuki Muramatsu (S&VL)･Makoto Iwamura (Fukuoka University)

This study develops a steering control method aimed at improving the cornering stability of vehicles equipped with steer-by-wire systems. By employing a hierarchical gradient method, which enables stable optimization under multiple constraints, the steering angle is derived to enhance cornering stability. This paper presents the application method of the hierarchical gradient approach to steering control and reports the initial evaluation results obtained using a driving simulator.

Study on Predictive Haptic Guidance Using Phase-Lead Compensation in Urgent Takeover of Automated Driving

Kansei Hattori･Toshio Ota (Hiroshima Institute of Technology)

In the urgent transition of control from automated driving systems to drivers, steering delays caused by cognitive lag present a critical challenge. This study proposes a system that performs predictive steering operations through phase-lead compensation and generates corresponding steering haptic feedback. Furthermore, this study investigates the effects of this predictive haptic guidance on the cognitive delays of drivers.

Ground Layout Design Considering Voltage Drop in Multiple Electronic Components

Yoshiko Seino･Masashi Komada･Takayoshi Shigihara (Toyota Motor)

The stable operation of multiple electronic components related to vehicle handling stability requires a ground layout that reduces voltage drop. In this study, resistance calculations are performed using a full-vehicle body shell model, and a method for designing ground layouts that reduces variations in inter-ground potential differences and voltage drops under combined current input conditions at multiple grounding points is presented.

Vehicle Lateral Control Based on Data-Driven ADRC

Shuichi Yahagi (Tokyo City University)･Itsuro Kajiwara (Hokkaido University)

This work introduces a data-driven design of the active disturbance rejection control (ADRC) for vehicle lateral control. This control method is effective for industrial systems with nonlinear and/or time-varying characteristics. However, its performance heavily depends on the choice of design parameters. The proposed method incorporates a one-shot data-driven design approach to optimize ADRC parameters directly from a set of data without repeated experiments. The effectiveness of the proposed method is demonstrated using a vehicle simulator, and the results confirm that it achieves good tracking performance.