Integrating Model-Based Observer and Kullback–Leibler Metric for Estimating and Detecting Road Traffic Congestion

Abstract

Efficient detection of traffic congestion plays an important role in the development of intelligent transportation systems by providing useful information for rapid decision making. The aim of this paper is to design an approach for road traffic congestion estimation and detection. Here, we design an innovative observer by integrating a hybrid piecewise switched linear (PWSL) traffic model with a Luenberger observer estimator for enhanced road traffic density estimation. This observer termed PWSL-LO combines the flexibility of the PWSL model with the simplicity and efficiency of a Luenberger observer to estimate the unmeasured traffic density. Moreover, this paper proposes an approach to monitor traffic congestion based on Kullback–Leibler distance (KLD) and exponential weighted moving average (EWMA) procedure. Residuals from the PWSL-LO model are used as the input for the KLD-EWMA scheme for congestion detection. This is motivated by the high capacity of KLD to quantitatively discriminate between two distributions. Here, the EWMA scheme is applied to the KLD measurements for congestion detection. Moreover, wavelet-based multiscale filter, a powerful feature/noise separation tool, is used to deal with the problem of measurement noise in the data. We evaluated the detection performance of this scheme by using traffic data from the four-lane SR-60 freeway in southern California. The proposed approach showed good abilities to estimate, monitor traffic congestions, and handle noisy traffic data.