Dual particle filter to identify headway distance of platooned vehicles

This paper applies a particle filter (PF) and a dual particle filter (DPF) for estimating the headway distance of vehicles traveling in a longitudinal platoon. Assuming that prior knowledge on the model parameter is different from what is observed in a real car-following, PF attempts to estimate the headway by measuring the acceleration rate and velocity of probe cars installed in the platoon. In addition, DPF is proposed to improve the accuracy by estimating the headway as well as the model parameters simultaneously within a single process. Numerical analyses were carried out to evaluate the performance of PF and DPF in two scenarios with six-vehicle platoon and different installation of probe cars. It was found that PF and DPF perform very well if all traveling vehicles in the platoon measure the acceleration rate and velocity. However, the effect of dual filtering process in DPF is not very significant to improve the estimation accuracy because the model parameters are not corrected appropriately by the dual filtering process.