Driver Steering Behavior Modeling Based On Visual And Muscle System

Driving control behavior of driver is one of most significant topics for the study of the entire closed-loop system,as the core element of the human-vehicle-road system.Therefore,after the investigation of driver models in domestic and foreign references,this paper proposes a driver steering model that combines a two-point preview vision model and a neuromuscular system model.It can simulate the situation where the driver turns to avoid obstacles,and it verifies the advantages of better tracking performance of two-point preview compared with single-point preview,as well as the influence of visual compensation system and the various parameters of the neuromuscular system on the vehicle state output.The following are the specific research contents:In the beginning,by establishing a vehicle-road system model,it will pave the way for the subsequent establishment of the overall model of the human-vehicle-road closed-loop system.Among them,vehicle modeling includes the establishment of vehicle kinematics model,dynamics model and steering system model.In order to simulate the driver's steering obstacle avoidance behavior,MATLAB is used to obtain the single-track line trajectory,and then input the horizontal and vertical coordinates of each point into CarSim to fit the simulated road.Furthermore,by analyzing the driver's visual attention mechanism during the steering process,a two-point preview mechanism is selected to establish a two-point preview model that includes visual prediction,visual compensation and signal transmission delay.Based on the mechanical characteristics of the neuromuscular system,the existing neuromuscular model is simplified,and a neuromuscular system model including feedforward control,feedback control and muscle dynamic control is established.As a result,the interaction between the visual system and the neuromuscular system is investigated through an overall analysis of steering behavior.By establishing a driver steering model that combines a two-point preview vision model and a neuromuscular model,it is combined with the vehicle-road model to form a closed-loop control system for human-vehicle-road.Afterward,it was verified by numerical simulation for the rationality of the model construction.