Regression Analysis of Heart Rate for Driving Fatigue using Box-Behnken Design

Abstract

There are few road accident studies that use heart rate as an indicator of driver fatigue. This study offers a mathematical regression analysis to discover which independent variables (driving speed, driving duration, body mass index, gender, and types of roads) are significant in influencing the heart rate and how these parameters interact to indicate driver fatigue. The analysis is conducted using a Box-Behnken design by Design Expert software. The results revealed that the values of Prob>F for all variables were less than 0.01%, indicating that all variables influence heart rate significantly. The heart rate increased when driving speed, driving duration, and body mass index (BMI) increased. A similar pattern was observed as the driving path shifted from urban to a moderately difficult uphill/downhill road. However, the pulse rate was reduced when a female driver was replaced by a male driver. The model's accuracy was evaluated by comparing the output data obtained from actual road driving with software prediction. First, the prediction interval of both techniques' output data was within 95%, meeting the minimum quantitative criteria of a 90% predictive interval. Subsequently, the residual errors were less than 10%. The regression model will be useful to shed light on traffic safety measures for preventing fatigue-related road accidents.