Matrix Estimation Pack

Multi-constraints matrix enhancement

Obtaining an accurate view of complex demand patterns is essential for validation of models of passenger travel and freight transport. OmniTRANS matrix estimation provides the tools to elicit information from observed travel data according to purpose, commodity, mode or any other category, enabling improved estimates of demand and reducing the scope and cost of data collection.

The quality and reliability of a model for use in a variety of planning applications depends strongly on the data available to reflect the activities and travel behaviour being modelled. Good trip matrices are of important influence on model outputs. OmniTRANS offers a modelling class to enhance travel pattern matrices using multiple types of constraints.

A model sufficiently robust to cover a wide range of applications however, can consume significant costs in conducting surveys and data collection. OmniTRANS offers a way to reduce this cost.

The core philosophy of Omnitrans International is to produce software that reduces the cost and effort of model development and allows model developers and planners to realise the full potential of their applications. In keeping with this philosophy, OmniTRANS offers the matrix estimation pack, which is designed to elicit maximum value from observed data in the specification and validation of models. And consistent with the design of OmniTRANS, the matrix estimation pack enables the user to unlock the hidden relationships between aggregate and disaggregate data, improving the specification of models for multiple purposes, modes, periods or any other dimension of planning.

Constraints

The matrix estimation class OtMatrixEstimation is designed such that the modeller can use observed data to an ultimate level. A variety of constraints is supported in the matrix enhancement process, including single traffic counts using detection loops, road-side interviews and licence plate detection. Also observations on productions and attractions can be used, as well as a priori matrices.

Estimation multiple matrices simultaneously

The unique concept of the OmniTRANS estimation class is that multiple matrices can be estimated simultaneously. For example when traffic counts on a highway are available in total, while road side interviews at a different location are known by mode of transport, both data sets can be used in the estimation process. Since each dataset can be linked to multiple matrices at once, the estimation process automatically uses these sets to calibrate the matrix contents.

Multi-routing

The enhancement process uses screenline intercept matrices generated by the static assignment class OtTraffic. These allow for multi-routing in the same way as simulated during a static assignment and can be generated for an unlimited number of constraints.

Dimensions and reliability

The matrix estimation class exploits the concept of dimensions ultimately. It can be easily embedded in a multi-modal and multi-temporal model framework where also other aspects like departure time choice can be used. Another interesting feature is the availability of applying weights to each constraint. This way you can assure that data from reliable count sites is used optimal, while less reliable count sites will not skew your trip matrices.

Feature diagram

The Matrix Estimation Pack gives access to the following features:

Requirement

This pack requires the presence of the Data Manager pack and the Static Assignment pack.