Investigating a model for predicting the traffic speed under rainy weather conditions using a neural network (Case study: Iranian province of Guilan)

Toanalyze more accurately traffic behavior, identifying relationships between major components of traffic characteristics is usually predominant. Although many research studies have been conducted based on the assumption that the weather would be sunny,the present study aims to propose a new model for predicting free flow speed utilizing neural network algorithm considering hourly rainy condition for a divided two-lane highway located in the Iranian province of Guilan. In this regard, a neural network algorithm was constructed using three input traffic data including hourly volume, heavy vehicle percent, and weather condition (1: rainy and 0: sunny). Theindependent variable of model is average travel speed. The results illustrate high accuracy according to the obtained correlation coefficient of 0.86.