Improving safety in mixed traffic: A learning-based model predictive control for autonomous and human-driven vehicle platooning

• DOI: 10.1016/j.knosys.2024.111673
• Publication Date: 2024-03-26T15:33:57Z
• Authors: Jie Wang; Zhihao Jiang; Yash Vardhan Pant
• Keywords: SAFER; Model Predictive Control; Baseline (sea)

As autonomous vehicles (AVs) become more common on public roads, their interaction with human-driven (HVs) in mixed traffic is inevitable. This requires new control strategies for AVs to handle the unpredictable nature of HVs. study focused safe mixed-vehicle platoons consisting both and HVs, particularly during longitudinal car-following scenarios. We introduce a novel model that combines conventional first-principles Gaussian process (GP) machine learning-based better predict HV behavior. Our results showed significant improvement predicting speed, 35.64% reduction root mean square error compared use alone. developed strategy called GP-MPC, which uses proposed safer distance management between platoon. The GP-MPC effectively utilizes capacity GP assess uncertainties, thereby significantly enhancing safety challenging scenarios, such as emergency braking In simulations, outperformed baseline MPC method, offering efficient vehicle movement traffic.