Investigating Factors Affecting Travel Time Using Multivariate Linear Regression Model with Robust Methods and FGLS

Articles in Press

Document Type : Research Paper

Authors

1 Associate Professor, Civil Engineering Faculty, K.N. Toosi University of Technology, Tehran, Iran.

2 Assistant Professor of Transportation Engineering, Tarbiat Modares University, Tehran, Iran.

3 PhD Student in Transportation Engineering, K.N. Toosi University of Technology, Tehran, Iran.

Abstract

This paper, using multivariate linear regression model with consolidated method and FGLS, investigates the factors affecting travel time between 113 traffic zones of Qazvin city with non-forced travel purposes at peak traffic time. The present article uses the extracted information from the comprehensive transportation studies of Qazvin. The model is used to pay for the least squares summarized, consolidated, and FGLS coding methods in the R-Studio software environment. Subsequently, the regression assumptions were examined and it was observed that the assumption of variance homogeneity was not established. Therefore, the consolidated method and FGLS payment method change were used and the results were compared. The results of these models include the increasing effects of variables on the use of public and private vehicles, personal vehicle ownership, and the existence of academic and commercial uses in travel destinations, and the diminishing impact of having high-level business and educational qualifications on time. Travel pointed. This article illustrates the impact of using consolidated methods and FGLS. Also the coefficient of determination of model fit with OLS and FGLS payment methods are 0.431 and 0.558, respectively.

Keywords

Experimental Research in Civil Engineering

Articles in Press, Accepted Manuscript
Available Online from 21 February 2022

Files

Share

How to cite

Statistics