To address the deterioration of freight vehicle control accuracy due to icy and snowy environments, an adaptive control method based on road friction coefficient estimation and drivers’ reinforcement theory learning is proposed. Road features are estimated without adding extra sensors because different roads have different tire slip rate-adhesion coefficient characteristic curve slopes within the range of tire linearity; the robustness of the control model is improved through research on the behavioral characteristics and parameters of experienced drivers on icy and snowy roads, design of a reinforcement learning reward function, use of a transverse and longitudinal control model training method based on deterministic strategy gradient reinforcement learning, and adversarial reinforcement learning. The scenario test of self-driving functions under icy and snowy environments carried out in Heihe city has demonstrated stable control of autonomous vehicles on ice- and snow-covered roads.
Adaptive Control Model for Unmanned Cargo Vehicles on Icy and Snowy Roads
2023-10-11
3198081 byte
Conference paper
Electronic Resource
English
Influence of Winter on Control of Snowy and Icy Roads in Inner Mongolia
British Library Conference Proceedings | 2003
|Influence of Winter on Control of Snowy and Icy Roads in Inner Mongolia
British Library Conference Proceedings | 2003
|Tires Adhesion Characteristic Analysis on Icy and Snowy Roads Based on Acceleration
Trans Tech Publications | 2011
|RUBBER COMPOSITION FOR TIRE FOR ICY AND SNOWY ROADS, AND STUDLESS TIRE
European Patent Office | 2018
|