摘要：We improve the typical GM car-following model. In order to describe the impacts of the lateral offset between the vehicles in car-following state, this paper proposes time distance to measure the drivers' attention of driving environment and uses generalized expectation headway to describe the ideal driving state of the driver. Take the difference between the actual and expected headway as the important standard for driver to decide whether to speed up or slow down. Then take the drivers' attentions and the expected headway as the decision-making factors, and consider them while revising the GM car-following model. The numerical simulation shows the improved GM model conforms to the general driving behavior, enhances the function of the typical model, and strengthens the model stability.
