Traffic safety is one of the most essential aspects of transportation engineering. However, most crash prediction models are statistically-based prediction methods, which require significant efforts in crash data collection and may not be applied in particular traffic environments due to the limitation of data sources. Traditional traffic conflict studies are mostly field-based studies depending on manual counting, which is also labor-intensive and oftentimes inaccurate. Nowadays, simulation tools are widely utilized in traffic conflict studies. However, there is not a surrogate indicator that is widely accepted in conflict studies. The primary objective of this research is to develop such a reliable surrogate measure for simulation-based conflict studies. An indicator named Aggregated Crash Propensity Index (ACPI) is proposed to address this void. A Probabilistic model named Crash Propensity Model (CPM) is developed to determine the crash probability of simulated conflicts by introducing probability density functions of reaction time and maximum braking rates. The CPM is able to generate the ACPI for three different conflict types: crossing, rear-end and lane change. A series of comparative and field-based analysis efforts are undertaken to evaluate the accuracy of the proposed metric. Intersections are simulated with the VISSIM micro simulation and the output is processed through SSAM to extract useful conflict data to be used as the entry into CPM model. In the comparative analysis, three studies are conducted to evaluate the safety effect of specific changes in intersection geometry and operations. The comparisons utilize the existing Highway Safety Manual (HSM) processes to determine whether ACPI can identify the same trends as those observed in the HSM. The ACPI outperforms time-to-collision-based indicators and tracks the values suggested by the HSM in terms of identifying the relative safety among various scenarios. In field-based analysis, the Spearman’s rank tests indicate that ACPI is able to identify the relative safety among traffic facilities/treatments. Moreover, ACPI-based prediction models are well fitted, suggesting its potential to be directly link to real crash. All efforts indicate that ACPI is a promising surrogate measure of safety for simulation-based studies.

Crash prediction models are used to estimate the number of crashes using a set of explanatory variables. The highway safety community has used modeling techniques to predict vehicle-to-vehicle crashes for decades. Specifically, generalized linear models (GLMs) are commonly used because they can model non-linear count data such as motor vehicle crashes. Regression models such as the Poisson, Zero-inflated Poisson (ZIP), and the Negative Binomial are commonly used to model crashes. Until recently very little research has been conducted on crash prediction modeling for pedestrian-motor vehicle crashes. This thesis considers several candidate crash prediction models using a variety of explanatory variables and regression functions. The goal of this thesis is to develop a pedestrian crash prediction model to contribute to the field of pedestrian safety prediction research. Additionally, the thesis contributes to the work done by the Federal Highway Administration to estimate pedestrian exposure in urban areas. The results of the crash prediction analyses indicate the pedestrian-vehicle crash model is similar to models from previous work. An analysis of two pedestrian volume estimation methods indicates that using a scaling technique will produce volume estimates highly correlated to observed volumes. The ratio of crash and exposure estimates gives a crash rate estimation that is useful for traffic engineers and transportation policy makers to evaluate pedestrian safety at signalized intersections in an urban environment. / Master of Science