Home
Description
People
News
Timeline
Papers

Updateable Traffic Simulations

Steve L. Ferenci

Richard M. Fujimoto

Simulation is widely used to assess the behavior of urban transportation systems for near- and long-term planning or as a decision aid to manage operational systems. A major issue is to evaluate how changes in design, operation, or incidents affect traffic flow. This usually requires performing a simulation study in which a family, or many families, of closely related simulations are performed. Completion of such studies can be very time consuming, potentially precluding the use of such simulation in operational plannig. To address this issue we have developed an approach called updateable simulation whose purpose is to reduce the time to complete multiple executions of a simulation. This technique updates the results of a prior simulation run rather than re-execute the entire simulation to take into account variations in the underlying simulation model or new inputs into the simulation. The approach uses execution logs of prior runs that will be used as a basis to derive additional simulation results. For example, one could model the system during peak loads in the absence of accidents. Then many simulations can be updated from the initial run to evaluate courses of action when incidents occur. By updating the simulation the time to perform a family of related simulations can be reduced. This research is developing updateable simulation methods and their application to transportation systems, and is examining realization in both sequential and parallel computing environments for both on-line and off-line applications.