Road Adhesion Coefficient Estimation Based on Vehicle-Road Coordination and Deep Learning

Accurate estimation of the road adhesion coefficient can help drivers and vehicles perceive changes in state effectively, reducing occurrence traffic crashes accordingly. Therefore, this paper proposes a method based on vehicle-road coordination deep learning. Firstly, vehicle-based data feedback system combined with network cloud is introduced, CarSim simulation used to expand set train model effectively. Then, dynamic analysis whole vehicle carried out, operation related are obtained as input model. Then self-attention (SA), convolutional neural (CNN), long short-term memory (LSTM) established, reduce instability prediction, Q-learning optimize weight Finally, verified by actual data. The results show that Ccloud effective, compared other commonly model, proposed effectively predict coefficient.