A Gradient Boosting Tree-Based Recognition Model for Low-Carbon Driving Behavior in Motor Vehicles

Current research shows that the low-carbon nature of motor vehicles has an important correlation with drivers’ driving behavior, and Low-carbon driving behaviors in motor vehicles have become a hot topic. Most of the existing studies focus on typical driving behaviors such as driving at a constant speed, normal acceleration, and deceleration, while this paper considers the short-term volatility of driving behaviors and introduces mixed driving behaviors to study short-term driving behaviors. Based on the gradient boosting tree algorithm, the motor vehicle low-carbon driving behavior recognition model is constructed, respectively, and the comprehensive performance of the model is better. Finally, based on the characterization of low-carbon driving behavior, this paper proposes low-carbon driving suggestions such as driving smoothly and reducing short-time state switching.