Quantifying drivers’ driving skills using closed-loop dynamic simulations of articulated heavy vehicles

Abstract

This paper presents a method for quantifying drivers’ driving skills using closed-loop dynamic simulations of articulated heavy vehicles (AHVs). Due to AHVs’ multi-unit configurations, large sizes, and high centers of gravity (CGs), these large vehicles exhibit poor directional performance. AHVs require wide roads and large radii of path curvature for evasive maneuvers; these large vehicles frequently display unstable motion modes, including trailer sway, jackknifing, and rollover. The directional performance of AHVs is frequently evaluated in terms of maneuverability and lateral stability. There exists a trade-off between the performance measures of maneuverability and lateral stability. An AHV, its driver and the road constitute a unique closed-loop dynamic system. The unique dynamic characteristics of AHVs impose significant challenges for safe operations of these vehicles. However, little attention has been paid to the interactions of driver-AHV. This paper tackles the problem of quantifying drivers’ driving skills considering the interactions of driver-AHV under simulated single lane-change (SLC) maneuvers. Based on two driver characteristic parameters, namely preview time (PT) and transport delay (TD), we propose two performance measures, i.e., path-following score (PFS) and combined stability score (CSS), as the indicators for assessing the driving skills of AHV drivers under the simulated maneuvers. The driver-AHV closed-loop dynamic simulation is implemented using the built-in driver model and virtual vehicles developed in TruckSim. The numerical simulation results for rearward amplification (RA) measures are validated using driver-in-the-loop (DIL) real-time simulation. The simulation identifies AHV drivers’ driving skills in terms of lateral stability, path-following, as well as balanced path-following and stability regions considering PT and TD properties.