EFFECT OF GEOMETRIC ELEMENTS AT CORNER ISLAND ON DRIVER BEHAVIOUR

Abstract

Road accident is a significant cause of death. Road junctions are prone to an accident because they must merge traffic. As per the Road Accident in India (2020 ) most of the traffic accident were occurred at various intersections. At the intersection, the key issues are conflict points where vehicle converge, limited visibility, and traffic congestion. Therefore to avoid such problem corner island are constructed. Factors such as traffic volume, geometry of islands, and driver demographic significantly influence driver behaviour at these locations Therefore, the present study examines driver behavior at the corner island. The main objective of this study is to invesyigate the effect of geometry of islands, traffic volume and driver demographic on speed, acceleration, and deceleration characteristic and also model of speed and acceleration/deceleration behaviour Data has been collected in Raipur and Naya Raipur city. For data collection, 13 intersections were selected, with a total of 27 corner islands. Each island has different shape, size, entry width of the island, exit width of the island, and turning radius. Twenty-eight participants aged between 20 to 30 years participated in this study. The eligibility was that they should have a valid license. Out of the 28 participants, 12 were female, and 16 were male. Some of them were novice drivers, and others were experienced. Data was collected with the help of the V-Box instrument, and video is recorded using a video camera. Two private modes were considered for this study: motorized two-wheeler and Cars. V-Box test suite software is used for data extraction for speed, acceleration/deceleration, distance, and duration. Avidemux software was used for extracting traffic volume count. To check the effect of geometric features of a corner island on vehicle speed, and acceleration/deceleration, z-test and one-way ANOVA test were performed. For the Z-test excel was used and for the ANOVA test, Minitab software was used. To develop a model of speed and acceleration/deceleration regression technique was used. Six types of regression were performed namely, linear, polynomial (quadratic and cubic), square root, multiple linear and ridge regression. To check the best fit model, the goodness-of-fit measures like R2, standard error of regression (s), and MAPE value were calculated, and also residuals plot and equity plot are plotted. It was observed that before reaching to truning point of island driver speed is relatively higher, and speed continues decreases until reaching the entry point of island. After that driver starts increasing their speed as they identify the appropriate gap for safe manoeuver. The findings of this study were speed, acceleration/deceleration were affected by geometric features of island such as area, radius of turning, entry width, exit width, length of turning, and internal angle of intersection. It was also found that speed, acceleration/deceleration were affected by traffic volume, gender and mode. Radius exhibit stronger positive correlations with speed, and acceleration/deceleration as compared to Area, Entry width, Exit width, and length of turning. Multiple linear regression was best fitted for predicting the speed, and acceleration/deceleration of a category of vehicle being considered.