Study on Driving Behavior Characteristics of Vehicles Through Curved Road Scenarios

This paper delves into the characteristics of driving behavior and analyzes the behavioral features of drivers in curve passing scenarios based on natural driving datasets from both open road tests and enclosed facility tests. This enhances the human-like driving capabilities of autonomous vehicles. By establishing a data acquisition system for real vehicles, the paper separately collects natural driving data from both open roads and enclosed facilities, extracting and analyzing the driving behavior features during curve passing. The main focus is on the coupling relationship between the vehicle's entry speed, maximum lateral acceleration, rate of change in lateral acceleration, and deceleration actions within the curve, and the characteristics of the curve being passed. Ultimately, this provides data support for designing personalized autonomous or driving assistance systems with human-like operational features.