PRIOR MATRICES (O-D TABLES) FOR MICROSIMULATION MODELLING
Microsimulation traffic modelling is probably the most exciting development in traffic planning - at last stakeholders can be shown anticipated changes to traffic conditions of new network and/or land use arrangements. This imposes stringent requirements that the model of existing conditions is 'right' - meaning that it meets the requirements of model calibration and validation, such as DMRB or FHWA or RTA, and, that the result looks realistic to local stakeholders.
A key element to achieving this is representative demand matrices. Most modelling packages have sophisticated tools that will take a prior matrix and, with the guidance of the modeller, shape it, eventually, into a good matrix that well fits the data. However, getting a good prior matrix is time consuming and can be data intensive; a poor prior matrix is easy to prepare but leads to extensive costs, delay and frustration, as the modeller endeavours to re-shape it to meet calibration requirements.
We have developed an approach that produces high quality prior matrices, making the estimation of final demand matrices much faster and less frustrating than is currently the case. Imagine how fast your estimation process would have been had your prior matrix been close to meeting to DMRB or FHWA criteria at the outset.
This process has cut weeks of modelling effort from estimation processes for large-scale models. By offering this service on a fixed fee basis and reducing schedules it provides clients with a straightforward way of managing project risk.
