Designers of innovative control systems for vehicles such as Adaptive Cruise Control (ACC) lack tools for assessing the impact of their product on driver/vehicle performance. This paper describes a methodology for predicting the occurrence and severity of stereotypical motion conflicts identified for a ACC system design and vehicle platform. The idea is to represent the ACC system and the vehicle’s response characteristics in a mechanistic manner while the conflict variables are considered per their probability distributions as seen in empirical data. The methodology presented includes: dynamic and regression models of vehicle and ACC systems, conflict scenarios, conflict measures, condition variables embedded in empirical databases, and the processing of results to express conflict occurrence as cumulative density functions. Six conflict scenarios have been observed in field tests, with one of these that addresses the overshoot in headway that may accrue when approaching a slower moving vehicle from long range used as an example. The condition variables such as the prevailing vehicle speeds, longitudinal grade and the road curvature are varied, as appropriate, in accordance with the probability distributions seen in actual highway service. Obtaining results for a single scenario requires Monte Carlo simulations. To decrease the computational cost, a regression equation is developed for use in efficiently predicting performance over the full matrix of operating condition variables. Computed results are shown and discussed in light of the design decisions that face manufacturers concerned with introducing ACC products.

This content is only available via PDF.
You do not currently have access to this content.