Driver model has been developed since 1950s and quite a lot of attention has been attracted on this aspect. Today there is a body of knowledge regarding driver mathematic model. Since driving behavior is highly complex, involving psychology and physiology factors, it is not easy to use a simple driver model to represent all characteristic features comprehensively, according to the utilization of driver model, some assumptions are necessarily proposed in order to tackle the specific problem easily.
Optimal Preview Acceleration Driver Model was put forward based on the Preview-Following Theory proposed by Prof. Guo in 1983. This driver model is quite simple and easy to understand, and accurate for great curvature path following. However, recent study has shown that more work is needed to improve the path following ability for sharp corner or urgent corner (higher frequency steering input). This paper will investigate and analyze the Optimal Preview Acceleration Driver Model in details, and propose an improvement based on this classical driver model to make it suitable for more specific conditions, especially for urgent cornering. Finally, the modified driver model will be verified to be more capable of realizing high frequency steering behavior than the previous driver model.