Doctor of Philosophy in Engineering
Civil and Environmental Engineering
Mohamed Kaseko, Committee Chair
First Committee Member
Second Committee Member
Hualiang (Harry) Teng
Third Committee Member
Graduate Faculty Representative
Number of Pages
The objective of this study was to develop a family of car-following models that address the shortcomings of car-following models developed by General Motors (GM) in the 1950s. The developed models consist of separate models for acceleration, deceleration, and steady-state responses for congested freeway traffic conditions. The study calibrated the models using individual vehicle trajectory data collected on a segment of Interstate 101 in Los Angeles, California. Furthermore, the study validated the models using individual vehicle trajectory data collected on a segment of Interstate 80 in Emeryville, California. The study used nonlinear regression with robust standard errors to estimate the model parameters and obtain the distribution of the model parameters across drivers and for different pairs of following vehicles. The stimulus response thresholds that delimit acceleration and deceleration responses were determined based on Signal Detection Theory.
The results indicate that average drivers' response time lag is significantly lower for deceleration response than for acceleration response. This is intuitive because deceleration response is generally related to safety, thus, drivers are expected to respond faster than for acceleration response. Acceleration is a response that is related to drivers' desire to attain maximum speeds which is the less urgent need than safety. Additionally, drivers' response to negative stimulus is sometimes further aided by activation of brake lights for a leading vehicle that is braking. For similar safety reasons, the results show that average stimulus threshold is significantly lower for deceleration response than acceleration response and with higher magnitudes of parameters for deceleration response than acceleration response.
The results also indicate that drivers' behavior is significantly different for different vehicle being driven and/or followed. The results show that automobiles traveling behind large trucks have both lower magnitudes of acceleration and deceleration responses than when traveling behind other automobiles. These are unexpected results and could be due to inability of automobile drivers to see beyond large trucks in front of them.
Overall, the results confirm the need for separating models for acceleration and deceleration responses and for different pairs of following vehicles because they impact drivers' behavior differently. However, both the driver response time lags and stimulus thresholds are likely to depend on speed and vehicle separation. This research simplified the models and determined the driver response time lags and stimulus thresholds independent of these factors.
Acceleration; Braking; Car-following models; Driver response time lags; Driver stimulus response thresholds; Driving habits; Freeway driving; General Motors (GM); Highway driving; Interstate 80 (I-80); Emeryville; California; Interstate 101 (I-101); Los Angeles; California; Signal detection theory; Stimulus-response car-following models; Traffic safety; Traffic simulation models
Siuhi, Saidi, "Parametric study of stimulus-response behavior incorporating vehicle heterogeneity in car-following models" (2009). UNLV Theses, Dissertations, Professional Papers, and Capstones. 152.