PhD Dissertation: Peter Nilsson Traffic Situation Management for Driving Automation of Articulated Heavy Road Transports, -From driver behaviour towards highway autopilot

Date
10 October 2017 10:15–13:00
Place
Chalmers, Hörsalsvägen 8, sal HB2, Gothenburg

Welcome to attend the public defence of Peter Nilsson's PhD dissertation, entitled "Traffic Situation Management for Driving Automation of Articulated Heavy Road Transports -From driver behaviour towards highway autopilot".

Faculty Opponent
Dr David Cole, Engineering Department, Cambridge University, UK

Graduate School
Machine and Vehicle Systems, Department of Applied Mechanics

A light lunch will be served after the dissertation VEAS facilities, Hörsalsvägen 8), please inform Sonja Laakso sonja.laakso [at] chalmers.se if you would like to attend.

Abstract
In this thesis traffic situation management for driving automation of long combination vehicles is discussed. The automation targets high-speed driving in multiple-lane, one-way roads. Traffic situation predictions, traffic situation manoeuvres and driving principles are studied specifically. Traffic situation predictions relates to the functions used to predict how an observed traffic situation will evolve in the future. Traffic situation manoeuvres relates to decision-making regarding driving principles and control on a tactical level of driving. The developed methods and principles assume the existence of vehicle environment sensing functionalities. Furthermore, the methods have been verified using motion platform driving simulator experiments and desktop simulations.

In the proposed methods for traffic situation predictions, models of the subject vehicle, driver, road and surrounding traffic have been formulated. These models capture both subject vehicle dynamics and predicted motion of surrounding traffic. Also, a unique driver steering model for articulated vehicles has been derived. Moreover, traffic situation predictions for multiple-lane one-way road driving has been derived by using driver steering and acceleration models in a closed loop with the subject vehicle model. Also, a second approach to calculate actuation trajectories has been developed and evaluated using a model predictive control framework including on-line optimisation. The derived traffic situation manoeuvres include maintain-lane, lane changes and non-evasive abort manoeuvres.

It is envisaged that studying the important characteristics of manual driving will give insight into how to design driving automation especially in regard to mixed traffic with both manually driven and automated vehicles. Driving principles for driving automation are derived by using back-to-back comparisons of manual and automated driving in simulator experiments. Driving principles for initiation and execution of lane-change manoeuvres with surrounding traffic as well as managing mandatory road exits and lane changes in dense traffic have been studied and some driving principles for automation have been derived.

Keywords: articulated heavy-vehicles, long combination vehicles, vehicle dynamics, vehicle model, driver model, driving automation, driving simulator, driving principles

 

Safer – Vehicle and Traffic Safety Centre

SAFER is the open innovation arena where researchers and expertise work together to create safe mobility. Our traffic safety approach covers people, vehicles and the infrastructure – and together we contribute to safer road transports and smarter, more sustainable cities.

Contact information

SAFER

Lindholmspiren 3
SE-417 56 Göteborg
Sweden

 +46 31-772 21 06
 safer [at] chalmers.se