MULTINOMIAL LOGIT MODEL FOR TRANSPORTATION PLANNING

Abstract

The medium sized cities in India are characterised by the availability of wide variety ,of modes of transport. No scientific analysis seems to have been done to determine the factors going into the choice of a particular mode . There is a necessity to develop an optimal intermodel mix for a given set of constraints in terms of land uses, various modes available, their energy efficiency and other related factors. Since the medium sized cities are developing very fastysuitable policies have to be framed for incentives and disincentives to various prevalent modes to get optimal transport mobility with minimal resource utilization. The scope of the study is therefore to develop a behavioral disaggregate, multinomial logit model (MNL) and to identify factors which go in for mode choice for a typical middle sized city. To determine the coefficients of the MNL a computer program has been developed by Peter Stopher. This computer program has been modified and debugged by H.K. Sinha for use for operating conditions in India. The data pertaining to the city of Patna as already collected for some other study has been utilized to check tie --applicability of the computer program. The various factors which go in for choice of a particular mode have been related to resultant (iv) Probabilities of the choice of particular mode of transport. A set of variations indicating the above relationships has been studied. Certain limitations of the MNL have also been mentioned.