Sensitivity analysis of an optimization-based trajectory planner for autonomous vehicles in urban environments

Jason Hardy; Mark Campbell; Isaac Miller; Brian Schimpf

doi:10.1117/12.802599

16 October 2008 Sensitivity analysis of an optimization-based trajectory planner for autonomous vehicles in urban environments

Jason Hardy, Mark Campbell, Isaac Miller, Brian Schimpf

Author Affiliations +

Proceedings Volume 7112, Unmanned/Unattended Sensors and Sensor Networks V; 711211 (2008) https://doi.org/10.1117/12.802599
Event: SPIE Security + Defence, 2008, Cardiff, Wales, United Kingdom

Abstract

The local path planner implemented on Cornell's 2007 DARPA Urban Challenge entry vehicle Skynet utilizes a novel mixture of discrete and continuous path planning steps to facilitate a safe, smooth, and human-like driving behavior. The planner first solves for a feasible path through the local obstacle map using a grid based search algorithm. The resulting path is then refined using a cost-based nonlinear optimization routine with both hard and soft constraints. The behavior of this optimization is influenced by tunable weighting parameters which govern the relative cost contributions assigned to different path characteristics. This paper studies the sensitivity of the vehicle's performance to these path planner weighting parameters using a data driven simulation based on logged data from the National Qualifying Event. The performance of the path planner in both the National Qualifying Event and in the Urban Challenge is also presented and analyzed.

Citation Download Citation

Jason Hardy, Mark Campbell, Isaac Miller, and Brian Schimpf "Sensitivity analysis of an optimization-based trajectory planner for autonomous vehicles in urban environments", Proc. SPIE 7112, Unmanned/Unattended Sensors and Sensor Networks V, 711211 (16 October 2008); https://doi.org/10.1117/12.802599

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available