Traffic State Estimation Using Traffic Measurement from the Opposing Lane—Error Analysis Based on Fluctuation of Input Data

In this study, we propose a method of estimating traffic states using the traffic data observed from the opposing lane. As input data, our proposed method requires (1) probe vehicle trajectory on the target section and (2) vehicle data observed by vehicles running in the opposing lane. These input data identify the cumulative vehicle counts along the trajectories of the probe vehicles and observation vehicles in the time-space of interest. We applied the variational theory based on the kinematic wave to estimate vehicle trajectories over the entire time-space using the identified cumulative vehicle counts as the boundary condition. Furthermore, we used the Clark approximation to analytically examine the sensitivity of the estimate to the stochastic fluctuations of the input data and measurements. Validation using a hypothetical network confirms that the proposed method reasonably reproduces the traffic states and that the Clark approximation can accurately estimate the sensitivity.