DYNAMIC ORIGIN-DESTINATION MATRIX ESTIMATION ON MOTORWAY NETWORKS

摘要: This thesis describes research into the problem of tracking time varying Entry-Exit (EE) travel demand from series observations, applied to small networks such as motorway corridors. To eliminate underspecification that was shown be inherent this a model referred specified. The objective is not only improve theoretical underpinning, but also expand practical applicability dynamic EE-estimation methods. commonly used restrictions entry flows are exactly known, traffic counts do involve internal links, and times relative length discretisation period relaxed by explicitely taking errors in volume observations account, allowing for use link counts, relaxing assumption dispersion. Taking point departure, second step involved deriving system properties following model, estimators its parameters, investigating how additional sources information may estimates. (A)