Evaluation of the Efficiency of Neural Networks and Statistical Models to Determine Daily Traffic Volume of the Suburban Roads of Mazandaran Province

Shariar Zargar,Sepideh Aldini,Seyed Hoseini

doi:10.12944/cwe.10.special-issue1.28

Abstract

Realizing the traffic volume at the present time is frequently one of the concerns that occupies the planners’ minds in transportation. Knowing the current volume plays an important role in reflecting the performance of transportation system in the future. Traffic studies are based on observations and interpretations of the current circumstances .Since the present observations cannot be represented for the future status, it should be predicted by means of determined conditions. Annual Average Daily Traffic is one the measure to be used for the traffic volume, which has been mentioned in the codes. The fixed or non-fixed automated counters serve to count this volume. In Iran, Road Maintenance & Transportation Organization is responsible to count daily through different ways. In the present study, the data collected from the selected axes of Mazandaran Province was utilized to make a predictive model for traffic volume. It is fitted by data, linear and logarithmic regression models and also neural network model.

Highlights

This Prediction of travelling rate was the primary data enjoying complexity in this present study
The aim of the present study is applying neural network in non-linear regression, which it is the best alternative for performing such function of MLP networks
In this study.We have suggested an artificial neural network and regression

Summary

Introduction

This Prediction of travelling rate was the primary data enjoying complexity in this present study. It has frequently been discussed on specifying how many travels occur in a particular region and condition whereas has been introduced new methods to predict it. Making model is the most important methods that are utilized by researchers’ model to discover the specific relationships embedded in the data. The directly predictive techniques are those attempt to predict frequency volume as a model; for instance, mathematic model and indirect methods predict the demand for travelling by calling upon 4-level models in the traffic solutions. The present study included 5 sections: the first section focused on intruding the study. Afterwards, the famous authors’ works and related books were reviewed

Objectives

Methods

Conclusion