Traffic Operation Status Research Based on Multi-source Data Fusion

Abstract

Paper summary generated by OpenAI: This paper presents a comprehensive study on traffic operation status utilizing multi-source data fusion, addressing the limitations of traditional single-detector traffic data in modern vehicle-road cloud environments. By integrating data from LiDAR, HD cameras, and V2X communication units, the authors develop a multi-source traffic data fusion model using wavelet neural networks and a genetic algorithm-optimized BP neural network. Additionally, a traffic status classification model is established based on the fuzzy C-mean clustering algorithm. The results demonstrate a fusion accuracy of 93% for the optimized BP neural network, highlighting the model’s effectiveness in supporting intersection signal control and traffic guidance optimization.