Abstract:Accurate forecasting of traffic flow is able to provide powerful traffic decision-making basis for intelligent transportation system, but the traffic data's complexity and fluctuation, as well as and the noise produced during collecting information and summarizing original data of traffic flow make the traffic flow forecasting result to have large error. This essay suggests a solution to the aforementioned issues and proposes a fully connected time-gated neural network based on wavelet reconstruction (WT-FCTGN). In order to eliminate the potential noise and strengthen the potential traffic trend in the data, we adopt the methods of wavelet reconstruction and periodic data introduction to preprocess the data. The model introduces fully connected time-series blocks to model the whole information including time sequence information and fluctuation information in the flow of traffic, and establishes the time gate block to comprehend the periodic characteristics of the flow of traffic and predict the flow of traffic. The performance of WT-FCTGN model is validated on the public PeMS data set. The experimental results show that WT-FCTGN model has higher accuracy, and its mean absolute error (MAE), mean absolute percentage error (MAPE) and root mean squared error(RMSE) are obviously lower than that of the other algorithms. The robust experimental results prove that WT-FCTGN model has good anti-noise ability.

A NETWORK TRAFFIC ANALYSIS AND FORECAST BASED ON WAVELET TECHNOLOGY

Prediction of Self-Similar Traffic and Its Application in Network Bandwidth Allocation

Self-similar traffic prediction based on Lyapunov exponent

Research on the Comparison of Time Series Models for Network Traffic Prediction

A New Multifractal Traffic Model Based on the Wavelet Transform

Modeling and Simulation of Self-Similar Traffic Based on FBM Model

A Wavelet-Based Method to Predict Internet Traffic

Forecasting internet traffic with wavelet method

Joint Network Traffic Forecast with ARIMA Models and Chaotic Models Based on Wavelet Analysis

The Multifractal Model Based on the Discrete Wavelet Transform

A New Multifractal Wavelet Model for Network Traffic

A Cauchy-Laplace Multifractal Wavelet Model for Network Redundant Traffic

A new multifractal network traffic model

Multifractal Traffic Modeling and Its Parameter Estimation

WAVELET THEORY AND ITS APPLICATION IN TRAFFIC INCIDENT DETECTION

WT-FCTGN: A Wavelet-Enhanced Fully Connected Time-Gated Neural Network for Complex Noisy Traffic Flow Modeling

The Study of Short-Term Traffic Flow Forecasting Based on Theory of Chaos

The Short-Term Predictability Of Traffic Flow

Survey of Research on Anomaly Detection of Wavelet Analysis Based on Network Traffic

An Analysis of the Combined Wavelet-Gm (1,1) Model for Network Traffic Forecasting

An online approach based on locally weighted learning for short-term traffic flow prediction.