Abstract:Traffic flow is chaotic due to nonstationary realistic factors, and revealing the internal nonlinear dynamics of chaotic data and making high-accuracy predictions is the key to traffic control and inducement. Given that high-quality phase space reconstruction is the foundation of predictive modeling. Firstly, an improved C-C method based on the fused norm search domain is proposed to address the issue that the C-C method in the phase space reconstruction algorithm does not meet the Euclidean metric accuracy and reduces the reconstruction quality when the infinite norm metric is used. Secondly, to address the problem of insufficient learning ability of traditional convolutional combinatorial modeling for complex phase space laws of chaotic traffic flow, the high-dimensional phase space features are extracted using the layer-by-layer pretraining mechanism of convolutional deep belief networks (CDBNs), and the temporal features are extracted by combining with long short-term memory (LSTM). Finally, an improved probabilistic dynamic reproduction-based genetic algorithm (PDRGA) is proposed to address the problem of the hybrid model falling into a local optimum when learning the phase space law. Experiments are conducted in three aspects: phase space reconstruction quality analysis, comparison of optimization algorithm convergence, and prediction model performance comparison. The experimentation with two data sets demonstrates that the improved C-C method combines the advantages of the high accuracy metric of the L2 norm with the low operational complexity of the infinite norm, achieving a balance between reconstruction quality and algorithm efficiency. The proposed PDRGA optimization algorithm is a lightweight improvement of the traditional genetic algorithm (GA) and solves the problem that the model tends to fall into a local optimum by optimizing the initial weights of CDBN. Meanwhile, the five error evaluation indexes of the proposed PDRGA-CDBN-LSTM hybrid model are lower than those of the baseline model, providing a new modeling idea for chaotic traffic flow prediction.

An Improved Phase Space Reconstruction Method-Based Hybrid Model for Chaotic Traffic Flow Prediction

Traffic Flow Predicting Of Chaos Time Series Using Support Vector Learning Mechanism For Fuzzy Rule-Based Modeling

Chaotic property analysis and prediction model study for traffic flow time series

Adaptive Volterra Prediction Model for Traffic Flow Based on Chaos

Fuzzy Hybrid Framework with Dynamic Weights for Short‐term Traffic Flow Prediction by Mining Spatio‐temporal Correlations

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

Short-term Traffic Volume Time Series Forecasting Based on Phase Space Reconstruction

The Improved Forecasting Model for Short-Term Traffic Flow Based on Phase Space Reconstruction

Short Term Traffic Flow Prediction Research Based On Chaotic Local Model

A Hybrid Deep Learning Based Traffic Flow Prediction Method and Its Understanding

Hybrid Extreme Learning for Reliable Short-Term Traffic Flow Forecasting

A Two-Stage Hybrid Extreme Learning Model for Short-Term Traffic Flow Forecasting

Multiple Sources and Multiple Measures Based Traffic Flow Prediction Using the Chaos Theory and Support Vector Regression Method

A hybrid deep learning model for urban expressway lane-level mixed traffic flow prediction

Dynamic multi-scale spatial-temporal graph convolutional network for traffic flow prediction

Short-time traffic flow prediction using chaos time series theory

GA-SVR Traffic Flow Prediction Based on Phase Space Reconstruction with Improved KNN Method

Chaotic Characteristic Analysis for Traffic Flow Series and Dfpsovf Prediction Model

Optimization and Performance Evaluation of Hybrid Deep Learning Models for Traffic Flow Prediction

A Hybrid Model for Forecasting Traffic Flow: Using Layerwise Structure and Markov Transition Matrix.

Exploring Spatial-Temporal Relations Via Deep Convolutional Neural Networks for Traffic Flow Prediction with Incomplete Data