Adaptive anticipatory network traffic control using iterative optimization with model bias correction

Anticipatory signal control in traffic networks adapts the signal timings with the aim of controlling the resulting (equilibrium) flow patterns in the network. This study investigates a method to support control decisions for successful applications in real traffic systems that operate repeatedly from day to day. A main bottleneck in designing the daily control scheme is the presence of model uncertainty. Conventional methods adopt a two-step procedure, iteratively updating parameter estimation and control optimization. Inconsistency arises due to the inevitable structural modelreality mismatch. This paper proposes an iterative optimizing control method to tackle this limitation and drive the traffic network towards the true optimal performance despite of model uncertainty. This Iterative Optimizing Control with Model Bias Correction (IOCMBC) corrects model bias using measurements and the resulting reality-tracking metamodel is updated for the subsequent control optimization. Theoretical analysis on matching between the IOCMBC optimal solution and the true optimum is presented. A local convergence analysis is also elaborated to investigate conditions required for a convergent scheme. One critical issue is the involvement of the sensitivity (Jacobian) information of the real route choice behavior with respect to signal control variables. To avoid performing additional perturbations, we introduce a measurement-based implementation method for estimating the operational Jacobian that is associated with the reality. Numerical tests in a small network verify the effectiveness of the proposed IOCMBC method in tackling model uncertainty, as well as a practical setting for regulating the reality-tracking convergence.