AN AUGMENTED LAGRANGIAN DUAL ALGORITHM FOR MULTIPLE VEHICLE STOCHASTIC USER EQUILIBRIUM PATH FLOW ESTIMATOR

摘要

In order to calculate multi-vehicle OD traffic demand from traffic counts under congested conditions, multiple vehicle stochastic user equilibrium path flow estimator is proposed. And augmented lagrangian dual algorithm (ALM) is given for multiple vehicle OD demand estimation. Link capacity, flow equilibrium and priori OD demand constraints are transformed into penalty function. The original constraint optimal problem is transformed into the non-constraint optimal problem. Then a simple projection algorithm is adopted for the non-constraint optimal problem. In a medium-size road network, The numerical examples show affectivity of model and algorithm.