A particle filter for freeway traffic estimation

This paper considers the traffic flow estimation problem for the purposes of on-line traffic prediction, mode detection and ramp-metering control. The solution to the estimation problem is given within the Bayesian recursive framework. A particle filter (PF) is developed based on a freeway traffic model with aggregated states and an observation model with aggregated variables. The freeway is considered as a network of components, each component representing a different section of the traffic network. The freeway traffic is modeled as a stochastic hybrid system, i.e. each traffic section possesses continuous and discrete states, interacting with states of neighbor sections. The state update step in the recursive Bayesian estimator is performed through sending and receiving functions describing propagation of perturbations from upstream to downstream, and from downstream to upstream sections. Measurements are received only on boundaries between some sections and averaged within regular or irregular time intervals. A particle filter is developed with measurement updates each time when a new measurement becomes available, and with possibly many state updates in between consecutive measurement updates. It provides an approximate but scalable solution to the difficult state estimation and prediction problem with limited, noisy observations. The filter performance is validated and evaluated by Monte Carlo simulation.