Research on Lane Change Strategy considering Driver's Personalized Driving Behavior

• Published in: 2021 International Conference on Artificial Intelligence and Electromechanical Automation (AIEA) pp. 317 - 321
• Main Authors: Wang, Yi, Deng, Bo, Ou, Yang, Li, Zhe, Fan, Jie
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.05.2021
• Subjects: Analytical models; Data models; driving bahavior; hidden Markov model; Hidden Markov models; lane change; Markov processes; Predictive models; Random processes; real-world data analysis; Statistical learning
• DOI: 10.1109/AIEA53260.2021.00074

In view of the problem that human-drivers in the side lanes are prone to enter the team during the driving process of internet vehicle fleets, this paper proposes a personalized driver lane change prediction model based on modern statistical learning theory. In this paper, a lane changing model considering driver characteristics is constructed by combining the Gaussian Mixture Model (GMM) and the Hidden Markov Model (HMM). This method solves the difficulty of describing the static distribution characteristics and dynamic random process in driver behavior. Finally, the lateral acceleration experiment is designed to collect vehicle acceleration data, and the validity of the model structure is verified by using natural driving data.