Extending DRAM Model: Theory-Practice Nexus

Abstract

The DRAM and EMPAL models of household and employment location and land use, respectively, have seen numerous applications by regional planning agencies and metropolitan planning organizations. One reason for this is that compared with other location and land use models, they are relatively easy to use. The theoretical structure of these models is that of aggregate multinomial logit. They are representations of choice probability means for household types and employment types, with their structures being derived from location surplus formulations. In operation the parameters of these models’ equations can most easily be estimated by use of a nonlinear programming formulation rather than by logit regression. Use of this technique allows a great simplification in the data requirements for calibration. In addition, this technique permits testing of alternative or auxiliary (to the standard model structures) variables in the model attractiveness formulations. The development of the models’ equation structures is described first. The problems and techniques of calibration-estimation of the equations’ parameters are then described. Finally, several experiments in model augmentation undertaken by agency staff as a part of ongoing model implementation efforts are described. These experiments demonstrate the flexibility of the structure for both application and experiment.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 16 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.