Modelling Trip Generation Using Adaptive Neuro-Fuzzy Inference System in Comparison with Traditional Multiple Linear Regression Approach

Sameer Abu-Eisheh (An-Najah National University & Universal Group, Palestine); Mohammad Irshaid (An-Najah National University-Nablus, Palestine)

Development of trip generation models has been conducted mainly using the traditional Multiple Linear Regression approach, which sometimes might not necessarily result in appropriate models, especially with existence of many interrelated and complex relationships among several related socioeconomic variables. This study investigates the feasibility of using a relatively new approach, the Adaptive Neuro-Fuzzy Inference System, and compares the results with those using the traditional approach. This is conducted by developing a home-based general trip generation model for one of the Palestinian urban areas. The comparison between the two methods outcome and the associated validation results is done using the R-squared, RMSE, and MAE measures. The Adaptive Neuro-Fuzzy Inference System was found to be a useful tool and a promising technique for modelling household trip generation, which is shown to outperform the traditional approach, with more accurate results and closer predictions to actual values. Further exploration of the new approach in transportation studies is recommended.

Conference: UKSim-AMSS 22nd International Conference on Computer Modelling and Simulation, UKSim2020

Published: Mar 25, 2020

DOI: 10.5013/IJSSST.a.21.02.17