Investigators

Daniel V. McGehee
The University of Iowa
National Advanced Driving Simulator
Tim Brown
The University of Iowa
National Advanced Driving Simulator
Chao Wang
The University of Iowa
Industrial and Systems Engineering

Final Report

View PDF|

Final Report Summary

View PDF|

Search

‚Äč

Project

Convoluted Gaussian Process (CGP): An Alternative to Facilitate Analysis and Predictions of Multiple DPMs under Several Driving Conditions Using Driving Simulators.

This project aims at modeling the interactions among different driving performance measures (DPMs), e.g., standard deviation of lateral position (SDLP) and driving speed, under various driving conditions. The hypothesis is that different DPMs interact with each other and the successful modeling of such interactions could greatly improve the prediction accuracy and reduce the variability of DPMs at untried driving conditions. The project would use driving simulators data to train and test the proposed DPM interaction model, where the DPM prediction accuracy would be evaluated and compared with various alternatives, e.g., generalized linear model, to validate the effectiveness of the proposed model

Supporting links:
Webinar
Dataset