Research on Prediction Model of Expressway Traffic Abnormal Events Based on Nonlinear Internet of Things

ABSTRACT Abnormal highway traffic events, such as traffic congestion, traffic accidents and road obstacles, not only affect traffic fluency but also have a negative impact on safety and the environment. In this paper, a prediction model of highway traffic anomaly events based on nonlinear Internet of Things (IoT) technology is proposed. The model integrates IoT devices and sensor networks to collect multi‐dimensional data such as traffic flow, speed, and vehicle distance in real‐time, providing data support for traffic event prediction. Using a nonlinear dynamic system modeling method, combined with complexity theory and graph neural network (GNN), the model can process complex spatiotemporal traffic data of expressways and accurately identify and predict the occurrence of abnormal traffic events. When analyzing the traffic flow of expressways, we observe that the traffic flow of a key section is 13.4 vehicles/min, and the speed is 28.0 km/h. The traffic congestion degree of this section is 76.0%, while the average speed of the other section is 71.2 km/h, and the traffic flow reaches 50.0%. During the testing of the traffic abnormal event prediction model, the prediction accuracy rate of this road section reached 96.7%, indicating that the analysis of real‐time data through nonlinear Internet of Things technology can significantly improve the prediction accuracy and traffic flow prediction ability. The model shows high sensitivity and accuracy in spatiotemporal data analysis and provides reliable data support for the management and optimization of abnormal traffic events.