Design and Development of a Virtual Autonomous System

2022-01-5072

10/03/2022

Event
Automotive Technical Papers
Authors Abstract
Content
An autonomous vehicle or a driverless vehicle is a ground vehicle that is capable of sensing its environment and moving safely with little to no human input. The present study aims to validate existing literature on vehicle following, emergency braking, lane changing, overtaking, and jerk trajectory with regard to autonomous vehicles. By using the jerk trajectory equation, we are able to optimize it by implementing a complete overtaking procedure. We are also able to compare the variation in both the dynamic and kinematic bicycle model for the same vehicle and traveling the same distance. A new kinematic bicycle model is built, which is more simple to construct.
This project is performed using MATLAB and Simulink for creating and modifying the system blocks. There has also been an integration of 3D visualization software such as VR sink and Unreal Engine. We construct the vehicles by simple transfer functions, ACC and PID as the main system blocks for vehicles 1, 2, and 3, respectively. We also construct different driving modes, which are assigned to vehicle 2 depending on the speeds of vehicles 1 and 2.
The comparison for both dynamic and kinematic bicycles has also been performed. We see nearly a 16.67% variation in the Y direction due to the variation in lateral forces between the two types of models. The jerk trajectory is validated by graphical representation of its lateral position, speed, and acceleration. The overtaking maneuver is visualized in Unreal Engine. The new kinematic bicycle models has been validated by comparing with the in-built model. The graphs indicate the accuracy of the model. Further work has to be done in creating a complete traffic-based autonomous system.
Meta TagsDetails
DOI
https://doi.org/10.4271/2022-01-5072
Pages
10
Citation
Narain, C., and Patil, S., "Design and Development of a Virtual Autonomous System," SAE Technical Paper 2022-01-5072, 2022, https://doi.org/10.4271/2022-01-5072.
Additional Details
Publisher
Published
Oct 3, 2022
Product Code
2022-01-5072
Content Type
Technical Paper
Language
English