الفهرس 
Autonomous Vehicle Lateral Control
Vehicle ModelingOnly 14 pages are availabe for public view
 |  | from | 183 |  |  |
| | | from | 183 | | |
Abstract
Autonomous vehicles can effectively reduce accidents and help minimize traffic jams, parking scarcity and the time wasted commuting by individuals. An autonomous vehicle can maneuver its routes in shorter time if it can navigate on the limits of handling. This requires accurate steering commands to be applied to the vehicle which can be generated through an accurate steer by wire system that is properly modeled along with the vehicle dynamics.
A mathematical model for autonomous vehicle dynamics is proposed to make use of the maximum attainable acceleration the tire can handle without losing traction. In addition to investigating the vehicle stability during maneuvers proposed by the higher-level control modules. The model integrates a steer by wire system with a simplified electric power train considering the handling dynamics of the vehicle. It takes into consideration forward, lateral, yaw and roll motions which is more exact and effective than simple bicycle planar models. The stability of the vehicle and its maneuverability are investigated. The handling limits of the vehicle are simulated using the g-g diagram under different maneuvers and tracks. Lateral and longitudinal accelerations, lateral velocity, slip angle, side forces and yaw motion are all investigated and the total acceleration is plotted on a g-g diagram.
The vehicle dynamics model is validated experimentally using an NI MyRio data acquisition system tested on an actual vehicle. A model predictive controller for a steer by wire system is developed to accommodate all the navigation conditions of the vehicle, and the changes in loading and commands provided by the perception and path planning system. The controller processes the commands given by these systems and provides a feedback with the recommended maneuver with respect to the vehicle handling limits provided by the g-g diagram. The controller optimizes the path decided by the decision-making systems and provides the fastest corner without losing control like skidding or roll over