Traffic flow prediction utilizes existing data and forecasting techniques to predict future traffic conditions, thereby providing valuable insights for urban traffic planning and control. This paper conducts visualized processing and comparative analysis of traffic flow prediction results using the Kalman filter model, LSTM model, and three types of graph convolutional neural networks. It demonstrates that deep learning-based traffic flow prediction can enhance prediction accuracy and provides improvement suggestions for further accuracy enhancement. Traffic flow prediction utilizes existing data and forecasting techniques to predict future traffic conditions, thereby providing valuable insights for urban traffic planning and control. With the development of deep learning, achieving more accurate prediction results has become possible. In this paper, the basic concepts of traffic flow prediction are reviewed, and commonly used methods for traffic flow prediction are introduced. By analyzing the application outcomes of the Kalman filter model, the LSTM model, and convolutional neural network models in traffic flow prediction, we demonstrated the advantages of deep learning methods in predictive analytics, with the Chebnet model achieving higher accuracy in forecasting. Nonetheless, challenges persist in traffic flow prediction, including issues with model generalization, data sparsity, and interpretability, necessitating further investigation and resolution.
Short-Term Traffic Flow Prediction Based on Deep Learning Models
2024-11-08
490595 byte
Conference paper
Electronic Resource
English
Deep Learning for Short-Term Traffic Flow Prediction
Online Contents | 2016
|Short-term traffic flow prediction using deep learning
SLUB | 2023
|Short-term traffic flow prediction method based on deep learning
European Patent Office | 2025
|Short-term traffic flow prediction method based on deep learning
European Patent Office | 2020
|