Modeling the Lane Contention Algorithm for Control in Intelligent Vehicular Systems

Abstract

Delegating the responsibility of safe driving on an autonomous vehicle needs the autonomous system to be extremely robust. In an attempt to design a robust guidance system for the autonomous vehicle we decided to go refer the most complex system designed which is “humans”. In humans we peculiarly observe that there exists an array of senses we possess that provide the required robustness and efficiency in chores that we perform. It clearly displays that in human beings the various senses we possess precisely complement each other to support a process. On this basis we propose an introduction to optimization of lane detection and tracking. Also the scalability of computation for smart vehicles such as auto park assist, cruise mode or auto lane keep assist continue to grow for driverless manoeuvre thus removing the possibility of manual or judgmental error in driving. We model a ground vehicle in a turning manoeuvre, such as a step lane-change where modeling process follows that of a vehicle undergoing translational as well as rotational motion. We shall be using this to design the control algorithm for executing a lane change maneuver under a safety set of rules using Dynamic Matrix Control which is a subset of the Model Predictive Control. This requires that we transform the vehicle lateral variable into the fixed coordinates. The application depends on recursively obtained sensor readings in a feedback loop mode for processing and deploying a corrective action tot exploits non-linear functions of the state and finds control inputs such as state of system, position, acceleration, peer movement to recursively estimate and improve the quality of resulting estimation for collision avoidance and target localization.