Model-Based Risk Map for Roadway Traffic Crashes; Rept. for 1 Jul 06-30- Jun 08

摘要

Visualization of traffic safety data that transforms spatial data into a visual form can help highway engineers and traffic safety officials to effectively analyze the data and make decisions on which roadways and road side features to improve by providing the spatial distribution of the data. However, research efforts in the visualization of traffic safety data, which are usually stored in a large and complex database, are quite limited because of methodological constraints (Miaou and Song, 2005bp; Miaou, Tandon and Song, 2005, Smith, Harkey and Harris, 2001). For example, there are only a few model-based maps that can account for the high variance of traffic crash estimates in low population areas, and at the same time clarify overall geographic trends and patterns. In addition, designers of roadways historically did not take into account the full range of driver characteristics, such as driver perception-response time, age differences, etc. (Dewar and Olson, 2002). One of the most important components of the roadway transportation is the human driver whose error is a factor in about 90% of the traffic crashes (Treat et al., 1977). Therefore, it is very important for highway engineers and traffic safety officials to identify and understand the basics of human factors as relevant to driving and traffic safety. The objective of the proposed project is aimed at developing a user-friendly geographic information system (GIS) that displays traffic crash data in Arkansas, estimated traffic risk based on the statistical model to be developed, and human factors in traffic accidents.