Development of a Model-Based Drive-By-Wire System for Level 4 Automated Vehicles

Eco-driving algorithms enabled by Vehicle to Everything (V2X) communications in Connected and Automated Vehicles (CAVs) can reduce energy consumption by generating energy-efficient speed trajectories. One of the key challenges from an implementation perspective is to be able to execute the desired s...

Full description

Saved in:
Bibliographic Details
Published in2023 IEEE International Systems Conference (SysCon) pp. 1 - 3
Main Authors Perrin, Jackson, Gankov, Stanislav, Downing, Walter, Rengarajan, Sankar, Hotz, Scott, Bhagdikar, Piyush, Alden, Joshua
Format Conference Proceeding
LanguageEnglish
Published IEEE 17.04.2023
Subjects
Online AccessGet full text

Cover

Loading…
More Information
Summary:Eco-driving algorithms enabled by Vehicle to Everything (V2X) communications in Connected and Automated Vehicles (CAVs) can reduce energy consumption by generating energy-efficient speed trajectories. One of the key challenges from an implementation perspective is to be able to execute the desired speed or acceleration commands closely. A SAE Level 4 CAV is being developed in conjunction with eco-driving to help address this issue. A custom drive-by-wire (DBW) system was built for a Honda Clarity plug-in hybrid electric vehicle (PHEV) to enable precise actuation of accelerator, brake, and steering. Transient vehicle data collected over a variety of drive cycles was synthesized to model pedal characteristics as a function of speed and target acceleration. The developed transfer functions enabled feed-forward control of the accelerator and brake pedals with a closed-loop controller handling the rest. Low speed steering maneuvers were performed to characterize the required steering torque to maintain a steady-state steering angle at different vehicle speeds. Advanced vehicle simulation software was used to understand vehicle dynamics and wheel angle response at different speeds. The development of a model based DBW system along with implementation details and results are presented.
ISSN:2472-9647
DOI:10.1109/SysCon53073.2023.10131104