Typical Road Condition Recognition Using Machine Learning and Vehicle Dynamic Response for Intelligent Driving

Road condition recognition is a critical technique for vehicle dynamics control and intelligent driving systems. This paper proposes a road condition recognition method based on CarSim simulation and machine learning, using vehicle dynamic responses. First, four typical road surfaces including asphalt, gravel, off-road, and sandy are modeled in CarSim, and vertical vibration signals of the wheels are collected as vehicle dynamic responses. Then, to identify the optimal model and verify the efficacy of the recognition method, several machine learning algorithms are employed and compared for road condition recognition. Experimental results demonstrate that the proposed method can effectively identify different road types with high classification accuracy. This study provides a feasible method for road condition perception, which also provides a meaningful reference and technical support for the practical implementation of intelligent vehicle systems.