Probabilistic framework for calibrating of highway geometric design models
An often overlooked aspect of highway geometric design is to the proper account for uncertainty in design models, input variables, and design parameters. This paper is a motivation to transform the current design models from deterministic to probabilistic. Due to specific aspects of highway geometric design, this transformation requires new theoretical understanding, interpretation, and development of underlying safety concepts. Reliability theory offers a widely acceptable framework for tracing the propagation of uncertainty through various design models. After defining a rejection domain for design outputs, the probability that a certain dimensioning decision of a road element leads to unacceptable driving conditions can be used as a design safety metric. The design safety metric tests the design of a road segment under a hypothetical group of driving situations that are generated based on the elementary knowledge of variability in design inputs. It is evident that the safety quantified in this research is different from the conventional definition of safety as being associated with collisions. In this research we strive to draw the line between the two safety definitions. Hopefully, with further development of various geometric design models, a link can be realized between. The literature includes several probabilistic analysis of dimensioning decision such as median width, horizontal curve radius, and roadway and railway intersection design. There is however an absence of research that extends these isolated studies to cover the entire spectrum of geometric design cases. In this research we attempted to formalize and integrate probabilistic analysis into standard design models. Several of the ideas and concepts demonstrated in this research were inspired by reliability-based structural design.
|Conference||4th International Gulf Conference on Roads|
Ismail, K, & Sayed, T. (Tarek). (2008). Probabilistic framework for calibrating of highway geometric design models. In Efficient Transportation and Pavement Systems: Characterization, Mechanisms, Simulation, and Modeling - Proceedings of the 4th International Gulf Conference on Roads (pp. 349–369).