Calibration of microscopic traffic model: cross entropy method and probability sensitivity analysis

Calibration and validation techniques pave the way towards the descriptive power of car-following models and their applicability for analyzing traffic flow. However, calibrating these models is never a trivial task because of the existing of unobservable parameters and erroneous traffic data. This contribution puts forward a new calibration framework of car-following models based on the Cross-Entropy Method and Probabilistic Sensitivity Analysis. Cross-Entropy Method is able to identify parameters of car-following models by formulating it as a stochastic optimization problem and to analyse the parameter estimations statistically while Probabilistic Sensitivity Analysis is used to identify the important parameters so as to reduce the complexity, data requirement and computational effort of the calibration process. Empirical results of calibration of intelligent driving model indicate the power of Cross-Entropy Method for searching global optimum for the case of synthetic data and next generation simulation datasets. Furthermore, adopting several termination criteria indicates better property of convergence of CEM than genetic algorithm