Vehicle Traveling Speed Prediction Based on LightGBM Algorithm

Abstract

Vehicle speed is an important indicator that affects the efficiency and safety of highway traffic, so realizing accurate prediction of vehicle speed on expressways can help reduce traffic accidents and improve the service level of intelligent traffic control. Aiming at the problem that the existing vehicle speed prediction model based on machine learning algorithm can not take into account the high computational accuracy, strong generalization ability and fast computation speed, the vehicle speed prediction model based on LightGBM algorithm is proposed. The model takes the ETC transaction data as the research object, uses the Light Gradient Boosting Machine (LightGBM) algorithm to establish the vehicle traveling speed prediction model with road features and vehicle traveling features in the feature library as inputs, and compares it with eight machine learning algorithms, including Decision Tree (DT), eXtreme Gradient Boosting (XGBoost), Gradient Boosting Decision Tree (GBDT), Ridge Regression (RR), Support Vector Regression (SVR), Random Forest (RF), Classification and Regression Tree (CART) and Back Propagation Neural Network (BPNN), and the results show that: the vehicle driving prediction model based on LightGBM has the best overall performance and can realize the fast prediction of vehicle driving speed with high prediction accuracy and strong generalization ability.