Analysis of queue length prediction from probe vehicles problem with bunch arrival headways

This paper discusses the real-time prediction of queue lengths from probe vehicles for the Bunch arrival headways at an isolated intersection for undersaturated conditions. The paper incorporates the bunching effect of the traffic into the evaluation of the accuracy of the predictions as a function of proportion of probe vehicles to entire vehicle population. Formulations are presented for predicting the expected queue length and its variance based on Negative Exponential and Bunched Exponential vehicle headways. Numerical results for both vehicle headway types are documented to show how prediction errors behave by the volume to capacity ratio and probe proportions. It is found that the Poisson arrivals generate conservative confidence intervals and demand higher probe proportions compared to Bunched Exponential headways at the same arrival rate and probe proportion.