Statistical evaluation of data requirement for ramp metering performance assessment

Ramp metering is known to be an effective freeway control measure that ensures the overall efficiency and safety of a highway system by regulating the inflow traffic on-ramps. Therefore, agencies are required to frequently assess the performance of their ramp meters. However, one major challenge for the agencies conducting ramp metering performance assessments is the lack of knowledge about data requirements. Data requirements consist of the information regarding the duration of data collection for accommodation time (the time needed for the users to get familiar with the change) and for the evaluation time. In this paper, a non-parametric statistic approach is proposed that is robust to the underlying distribution of the random variable. Meaning that the accuracy of the model is insensitive to the data distribution. For validation purposes, three active ramps along State Route 51, in the Phoenix Metropolitan area, Arizona are selected as the case study. ADOT altered their ramp control strategy from fixed-time to responsive control and is attempting to know the extent of data required for assessing its new ramp metering strategy. For this particular case study, the results suggest that two months’ worth of data is the minimum data sufficient for a ramp metering assessment. The proposed assessment approach can be transferred to other ramp metering applications to help traffic engineers efficiently tune up ramp metering strategies.