(200b) A Control and Optimization Framework for Sustainable Energy and Transport Policies in Urban Environments | AIChE

(200b) A Control and Optimization Framework for Sustainable Energy and Transport Policies in Urban Environments

Authors 

Shukla, A. - Presenter, Purdue University
Peeta, S. - Presenter, Purdue University


As more and more people all over the world move from rural to urban areas, the travel demand on the transportation infrastructure of these population hubs is growing rapidly. This often leads to traffic congestion especially during the so called ?rush hour'. Urban Congestion not only leads to delays and unreliable travel times but also increased fuel consumption and emissions [1].

To this end metropolitan planning agencies need modeling capabilities to make informed decisions by testing the effect of traffic management policies before they are implemented on the system. This calls for the development of an integrated simulation-optimization based framework which can be used for evaluating the effects of traffic management policies. In this regard as a first step we need to simulate the transportation system to a level of detail which makes it responsive to policy changes, by capturing behavioral shifts in travel patterns over time. This will then be coupled with a feedback decision engine which optimizes policy parameters.

The history of transportation modeling and forecasting has been dominated by a modeling approach that has come to be known as the ?four step model' [2], where travel derived from activity participation by individuals, in practice is simulated with trip-based rather than activity-based models. Over the years as computational power became cheaper a new class of models called Multi Agent models [3] which use daily activities of the individual as the basis of travel demand [4] are becoming more popular. These multi-agent activity based models are now being tested on a few cities in Europe and North America. Also traditional Static assignment simulation models often used with the four-Step model are being increasingly replaced by  dynamic traffic assignment (DTA) models [5] which lead to more realistic demand assignment modeling with time-varying travel demand loads.

Our work intends to use Activity based multi-agent DTA models for travel demand forecasting in a feedback loop, to iteratively determine policy parameters until the desired policy set point is achieved. This state-of-the art modeling technique will be responsive to short and long term behavior changes after policy implementation in the test network, and will therefore be useful for making informed policy decisions. The methodology as visualized in process control framework is shown in Fig 1. Results from Preliminary work done with a Dynamic traffic Assignment scheme are now being integrated with multi-agent travel demand generator for better decision making.

References

1.            Schrank, D. and T. Lomax, The 2007 Urban Mobility Report. 2007, Texas Transportation Institute, Texas A&M University.

2.            McNally, M.G., The Four Step Model, in Handbook of Transport Modelling, D.A. Hensher and K.J. Button, Editors. 2000, Pergamon. p. 35-52.

3.            Balmer, M., K. Axhausen, and K. Nagel, Agent-Based Demand-Modeling Framework for Large-Scale Microsimulations. Transportation Research Record: Journal of the Transportation Research Board, 2006, p. 125-134.

4.            McNally, M.G., The Activity-Based Approach, in Handbook of Transport Modelling, D.A. Hensher and K.J. Button, Editors. 2000, Pergamon. p. 35-52.

5.            Peeta, S. and A. Ziliaskopoulos, Foundations of Dynamic Traffic Assignment: The Past, the Present and the Future. Networks and Spatial Economics, 2001, p. 233-265.