Calibration the Gravity Model for Areas with Limited Data by Maximum Likelihood Method : Case Study in Nairobi (KENYA)

Abstract

This research was concerned with the task of calibrating a model that can be used to map out travel demand in metropolitan regions in developing countries where the cost of data collection is not negligible. The Maximum Likelihood (ML) method is found to be an appropriate approach using traffic count data at road sections, and population and employment data for the zones. It calibrates the gravity model for OD matrix estimation without any constraints. The proposed model is then applied to Nairobi (Kenya) and it is found that the feasibility of the model is sufficient for estimating the OD travel demand. Assuming that the calibrated parameters remain stable, the model can be used to predict future travel demand if the land use patterns are known.