Traffic Assignment by Paired Alternative Segments: A Case Study based on an Ohio Travel Demand Model

摘要: This paper examines the performance of a new traffic assignment algorithm, known as TAPAS, in case study drawn from transportation planning practice. By storing and operating on paired alternative segments, TAPAS is able to converge quickly user equilibrium (UE) while finding most likely UE route flows, which satisfy so-called proportionality condition. The results show that can achieve more precise solutions compared commercial tool based well-known Frank-Wolfe algorithm. Further experiments involve comparison with research code confirm not only generates solutions, but also them much less computational effort. reveals precision levels do make difference aggregated analyses. Importantly, project benefit estimated by about 4% lower than given software, seems large enough influence decision making. authors note solution produced software this has achieved relative gap 1e-4, level considered many sufficient for scenario comparison. provides evidence higher may be necessary certain