Comparison of optimization methods for assisted calibration of traffic micro-simulation

Usage of traffic simulation has increased significantly over the past two decades; and this high-fidelity modelling, along with moving vehicle animation, has allowed important transportation decisions to be made with better confidence. During this time, traffic engineers have typically been encouraged to embrace the process of calibration, in which steps are taken to reconcile simulated and field-observed traffic performance. According to international surveys, top experts, and conventional wisdom, existing (non-automated) methods of calibration have been difficult and/or inadequate. There has been a significant amount of research on techniques to improve calibration, but many of these projects and papers have not provided the level of flexibility and practicality typically required by real-world engineers. With this in mind, a patent-pending (US 61/859,819) architecture for software-assisted calibration was developed to maximize practicality, flexibility, and ease-of-use. This architecture is called SASCO (i.e. Sensitivity Analysis, Self-Calibration, and Optimization). The original optimization method within SASCO was based on "directed brute force" (DBF) searching; performing exhaustive evaluation of alternatives in a discrete, user-defined search space. Simultaneous Perturbation Stochastic Approximation (SPSA) has also gained favor as an efficient method for optimizing computationally expensive, "black-box" traffic simulations, and was also evaluated within SASCO. Preliminary experiments were performed to compare the effectiveness of DBF and SPSA, using synthetic and real-world networks. Results imply the two optimization methods have complementary attributes, and in some cases should be applied in tandem. Regardless of which optimization method is selected, the SASCO architecture appears to offer a new and practice-ready level of calibration efficiency.