(A) study on a driver model for vehicle lateral dynamics

Abstract

The road-vehicle-driver models are required to be mathematically mapped into computer for the virtual driving simulation. Even though those are essential elements, lack of realistic driver model has been pointed as a problem with the driving simulation. There have been several studies to develop a realistic driver model in accordance with the increased interests in vehicle safety issues and computer simulation for a vehicle design. This study is especially considering the human drivers’ steering process; path planning, feed-forward steering, and preview feedback steering. Important human factors\mdashview angle and neuromuscular system\mdashare also regarded. In overview, the weighting strategy utilizing multi-preview points is considered to classify the skilled level. Target path is obtained in analogous to the planar motion of a rigid rod under holonomic constraints. Feed-forward steering model is based on the vehicle-road geometric conditions. A PID type preview control scheme is considered for feedback steering. In consideration of human factors, the driver’s view angle is considered in feed-forward steering, and the neuromuscular system dynamics is only included in feedback steering. The suggested driver model is simulated with CarSim Vehicle model in Simulink environment. The simulation results are compared to the actual vehicle test data and also to the driving simulator test data with regards to 4 levels of human drivers. The driver model suggested in this study represents the human steering behavior and also matches well with the real vehicle test data.