Trajectory Planning and Maneuver Control to Assist Lane Change

Abstract

This paper explores the way to maneuver the autonomous vehicle from one lane to another. In the advent of automating the ground vehicle, lateral control pays a pivotal role in determining the navigation of vehicles. The lane change feature is designed to aid drivers during maneuver from one lane to the adjacent lane. It involves a high-level interaction when an autonomous vehicle steer from one lane to another. The lane change is a standout amongst the most altogether researched programmed driving tasks that can be utilized by a self-driving vehicle. Many the exploration has been done beforehand to carry out lane change in driverless cars with the help of machine vision and complex controllers. In this work, the PD controller is used, and to check the favorable conditions, sensor topology is studied. To deal with dynamics, the bicycle model and constant acceleration models are verified and simulated. This trajectory has been designed using a polynomial equation method to increase the reliability of results which is efficient than other conventional methods. The inventive steps are illustrated in adopting polynomial-based path planning with constraints of vehicle dynamics and further aided with integrated lateral position control with predictive heading control. The proposed lateral control is an illustration of predictive motion control with weighted steer profiling considering non-linear vehicle dynamics to track planned path during automated maneuver. The performance of the lane change maneuver has been verified in simulation-based environment with the help of Simulink model and Carmaker vehicle dynamics in loop.