Pedestrian multi-class speed-density relationship: evaluation of integrated and sequential approach

The focus of this paper is the modeling of speed-density relationship for pedestrian movements using the potential of detailed tracking observations (Alahi et al., 2014). This relationship is in the literature usually specified under the assumption that the traffic system is at equilibrium, that is stationary and homogenous (Jabari et al., 2014). The empirical analysis we have performed rules out the use of a unique equilibrium relationship, due to a high scatter in the data. We attribute the observed scatter to the violation of the homogeneity assumption, and specify a multi-class model of the speed-density relationship. We propose the integrated approach and more data-driven sequential approach to the specification of the model, and evaluate their performance using real data.