Real-time classification of road surface for the electric vehicle

The road condition greatly influence vehicle dynamics control. The friction coefficient of road surface is the primary factor of the ABS/ASR systems. In this paper, a real-time classification of road surface based on the distributed driven electric vehicle is proposed. The Pacejka tire model and Burckhardt tire model are chose. The algorithm based on the single feed-forward neural network with extreme learning machine is used for road classification. The simulation and experiment test under a wide range of road conditions are conducted to confirm the effectiveness of the proposed method. The research results show that the algorithm is able to identify four surfaces: asphalt, wet, snow and ice road. The proposed approach has the ability to provide with reliable information for vehicle passive-active safety control.