A Speed Prediction Model on Horizontal Curves of Two-Lane National Highway: A case study of Nagdhunga-Naubise Road

Abstract

Speed prediction models has been used in many countries as a useful alternative in road design concept. Each model is different in terms of methodology, data accuracy, variability in highway geometry and predictor variables used to predict the operating speed. This research focuses on developing a relationship between vehicles operating speed as dependent variables and road geometric elements such as radius, deflection angle, and length of curves, carriage width, gradient, shoulder width and super elevation as predictor variables in Statistical Software (SPSS V20). The speed of more than 3000 free-flowing vehicles were measured along 37 horizontal curves. Three models of predicted 85th percentile curve speed with the effects of geometric elements were developed using multiple regression method. Radius of curve was found to be the most significant predictor. The proposed model was validated with the speed and road geometrical data of 10 other curves. The R2 values for the final predicted models at the start, mid and end of curves were obtained as 0.703, 0.643 and 0.626 respectively. Similarly, the RMSE values for start, mid and end of curves were obtained as 5.15km/hr., 2.97km/hr. and 8.48km/hr respectively whereas MAPE values for the same were obtained as 0.12, 0.09 and 0.15 respectively.