摘要: |
To assess the quality of data provided by traffic detectors, and therefore detector health, research is necessary to bridge the gap between existing traffic theory and pattern recognition that can identify poor performance (through comparison of data to expected norms) in an automated fashion. To address this gap, the proposed objective is to develop a reliable and robust method of determining poor performance of a traffic detector based solely on historical data and traffic flow theory. It is proposed that this method will work at isolated signalized intersections, using data only from that intersection’s detectors for evaluation. Additionally, a system design of this method will be developed to assist Oregon Department of Transportation (ODOT) with implementation of the method. |