01661262

Component

Cognitive ability and decision-making process in the driving task vary among drivers. Individual differences on driver’s judgment as well as sensitivity to operational factors such as front gap and velocity differential affect driver’s decisions in driving. Based on fuzzy logic, which involves uncertainty and linguistic terms, this study develops a multi-dimensional decision-making model with driving personalities. Fuzzy sets are created to involve factors affecting driver’s car-following, lane-changing, cut-in and filtering decisions. Different from traditional decision-making models, 3 critical parameters are defined to represent drivers’ level of perception, sensitivity and risk-taking attitudes. The proposed model is then applied to recognize driving personalities of 30 drivers from Shanghai Naturalistic Driving Study. Critical parameters for each driver is calibrated to extract each driver’s driving personality. These critical parameters are then clustered in 6 groups. Each group’s driving styles are identified through analyses of their centroid, namely Neutral; Cautious; Aggressive; High Speed; Risk-affine and Risk-averse are identified. Safety performances of each driving style are evaluated under various traffic conditions through a car-following experiment. It is found in this study that when driver’s driving personalities are extracted from multiple levels, a driver’s perception and sensitivity level towards factors such as gap and velocity can be very different. A driver can be significantly less sensitive to a certain factor while he or she is normal towards other factors. The proposed model is found to be able to serve to better understand different driving styles and to provide long-term values of designing customized Advanced Driver Assistant System (ADAS) system according to user’s driving style.

This paper was sponsored by TRB committee AND30 Standing Committee on Simulation and Measurement of Vehicle and Operator Performance. Alternative Title: Fuzzy Logic–Based Multidimensional Driving Decision-Making Model Involving Driving Styles Recognized from Naturalistic Driving Study

18-03668

English

4p

2018

Transportation Research Board 97th Annual Meeting

Digital/other

Figures; Tables

Car following; Cognition; Decision making; Driver performance; Fuzzy logic; Gap acceptance; Lane changing

Highways; Safety and Human Factors

Transportation Research Board Annual Meeting 2018 Paper #18-03668

TRIS, TRB, ATRI

Jan 8 2018 10:54AM