Modeling and Implementation of Rack- Electric Power Assisted Steering System for driving simulator

Abstract

Driving simulators play an important role in improving driver safety and supporting the development and testing of driver-assistance systems. Their usefulness depends on the ability of the vehicle model to reproduce the behavior of a real vehicle. Since vehicle response is governed by the mathematical model with interconnected subsystems (Powertrain, steering, etc), modeling each is necessary. This thesis focuses on the development of an electric power-assisted steering (EPAS) model. The developed steering subsystem was integrated into the vehicle model containing an electric powertrain. During vehicle testing, access to controller area network (CAN) signals was restricted, preventing direct logging of steering-related signals. Consequently, the steering-wheel angle was estimated from two inertial measurement unit (IMU) setup in the vehicle and subsequently used as an input to the steering model. The vehicle model was evaluated through objective and subjective validation. First, the simulated response with experimental inputs was compared with vehicle response using the ISO/TS 18571 objective rating method. The results showed that the developed model captured the main response trends but did not achieve perfect agreement with the experimental data. This was expected, as several vehicle subsystems outside the steering model were retained from the existing baseline vehicle model and were not re-parameterised for the electric vehicle (used for vehicle tests). Second, the vehicle model was integrated in the driving simulator andevaluated through a subjective study involving test participants. The developed electric vehicle model with EPAS was compared with a baseline model equipped with hydraulic power-assisted steering (HPAS). Since the HPAS model was used as a reference during EPAS model tuning, the subjective ratings did not show large differences between the two steering systems. The participants did not report a clear or consistent perceived difference between the models. The subjective feedback nevertheless provided valuable input for identifying areas for further refinement of the steering model. The developed model provides a basis for representing electric-vehicle steering architecture.