A Comparative Study of the Effect of Parallel vs Ackerman Steering on 14 DoF Real-Time Vehicle Simulation Model

The advent of autonomous driving features has brought about the rise of new automotive technologies such as drive-by-wire. But the implementation of these technologies on physical vehicles is more complex, creating a need for simulation, model-based development, and testing of these systems before they are implemented in an actual vehicle. The real-time simulations capabilities of MATLAB and Simulink provide a robust development platform for behavioral cloning. The most common high-fidelity model used for simulation of vehicle dynamics is the 14 DoF model. Ackerman steering geometry has been around for two centuries and it is the most commonly used steering geometry for passenger cars. Yet, 14 DoF vehicle simulation models in literature have been observed to be using parallel steering geometry due to their simplicity. This paper focuses on developing a 14 DoF model with Ackerman steering geometry and presents a comparative study of Ackerman and Parallel steering geometries in 14 DoF vehicle models which are validated against real vehicle test data. Another common assumption is that the slip angles of the left and right wheels are the same. This assumption inaccurately estimates the effect of the yaw velocity of the vehicle. The effect of this estimation is also examined in this research. Within the scope of this research, a 4 DoF tire model using Pacejka magic formula with combined slip, a 3 DoF handling model with Ackerman steering geometry, and a 7 DoF ride dynamics model which takes in the road profile as input is developed. The current research presents a Simulink model set up to take real-time inputs using open-source steering, brake, and throttle input hardware. The model is validated against real-world vehicle data and then the effects of the estimations are observed. This research presents the observations on vehicle yaw velocity, lateral acceleration, and slip angles and their effects on the overall performance.