Abstract:Obtaining future traffic state evolution information is critical to traffic control algorithms design and further to intelligent transportation systems. However, accurately predicting traffic state evolution is not an easy task, although the traffic prediction-related study attracted a lot of attention. This study develops a macroscopic traffic flow model-integrated deep learning framework ( $\mathbf{MTFD}$ ) for the high-resolution temporal-spatial traffic state dynamic propagation on a road network, integrating temporal-spatial traffic dependency, traffic flow theory, and data analysis techniques. First, traffic state propagation on every road section is mathematically described by the $\mathbf{CTM}$ model given traffic initial and boundary conditions. Next, a temporal-spatial traffic dependency attention ( $\mathbf{TSTD}$ ) recurrent neural network is developed to predict boundary conditions factoring the traffic temporal-spatial dependency. Also, this paper develops a graph theory-based method to capture the temporal-spatial traffic dependency among the traffic on neighboring road sections. Last, the extended Kalman Filter ( $\mathbf{EKF}$ ) is introduced to adjust the predicted traffic state at an intersection to satisfy the conservation law. The numerical experiments illustrate that the proposed method predicts the traffic state evolution in a freeway network within 30 minutes with accuracy varying from 75%-95%. It has a better performance compared to the tested baseline models (APTN, Graph CNN-LSTM, and so on). The experimental results also illustrate that factoring traffic dependency and integrating data assimilation techniques can improve prediction accuracy.

Multi-step traffic flow prediction model based on wavelet and echo state network

An Integrated Model for Traffic Flow Prediction Based on the Wavelet Transform

Research on Real-time Prediction Model of Road Traffic Flow based on Wavelet Neural Network

Traffic flow prediction based on wavelet transform and Radial Basis Function network

Short-Term Traffic Flow Prediction With Wavelet and Multi-Dimensional Taylor Network Model

The Traffic Flow Prediction Based on Wavelet De-noising and Chaotic Time Series Theory

Traffic Prediction considering Multiple Levels of Spatial-temporal Information: A Multi-scale Graph Wavelet-based Approach

Research on the Real-Time Prediction Model of the Traffic Flow Based on Wavelet Neural Network

Short-term traffic flow prediction based on improved wavelet neural network

Multistep Prediction of Flow Speed on Consecutive Sections of Expressways Based on State Space Model

A Hybrid Approach of Traffic Volume Forecasting Based on Wavelet Transform, Neural Network and Markov Model

Wavelet Transform Based Traffic Flow Predicting and Model Optimization

Research of Prediction Scheme Based on Adaptive Particle Swarm Wavelet Neural Network Model

Traffic flow cycle prediction based on the PCA-ESN model

Network-Wide Traffic Flow Dynamics Prediction Leveraging Macroscopic Traffic Flow Model and Deep Neural Networks

Multi-step Network Traffic Prediction Using Echo State Network with a Selective Error Compensation Strategy

Wavelet-based Short-Term Forecasting with Improved Threshold Recognition for Urban Expressway Traffic Conditions

Prediction of traffic flow with wavelet network based on improved genetic algorithm

Short-time traffic flow prediction using chaos time series theory

Multi-step Traffic Flow Prediction Method Based on the Conv1D + LSTM.

Analysis and prediction of complex dynamical characteristics of short-term traffic flow