An expert system for risk assessment of automobile crashes

摘要

There is an ongoing concern to make Alabama roadways a safer place to travel, The purpose of this research is to help in this effort. Identification of critical accident characteristics based on previous accident data helps to identify areas where safety efforts should be directed. Early recognition of critical crash characteristics represents the most efficient way to reduce the occurrence of traffic-related accidents. However, a major obstacle to the assessment of risk is the inability to provide necessary information concerning accident analysis. The objective of this research was to develop a computer-based expert system for risk assessment of automobile accidents. The Expert System for Risk Assessment of Automobile Crashes (ExRAAC) assesses accident characteristics based on inputs by the user. The system emulates the thought process of an expert, in the field of transportation engineering, in assessing the result of such an automobile accident. This expertise is represented in the knowledge base of the system using If-Then rules. The expert system asks the user questions regarding the characteristics of the accident, and based on the answers, assesses whether the accident will result in a 'fatality', 'injury' or 'no injury no fatality'. It also gives the probability of the occurrence of such an accident. The critical characteristics of the accident can then be further analysed to discover areas where safety efforts should be directed in order to prevent such a crash.