Wavelet-Based Denoising for Traffic Volume Time Series Forecasting with Self-Organizing Neural Networks

Abstract

In their goal to effectively manage the use of existing infrastructures, intelligent transportation systems require precise forecasting of near‐term traffic volumes to feed real‐time analytical models and traffic surveillance tools that alert of network links reaching their capacity. This article proposes a new methodological approach for short‐term predictions of time series of volume data at isolated cross sections. The originality in the computational modeling stems from the fit of threshold values used in the stationary wavelet‐based denoising process applied on the time series, and from the determination of patterns that characterize the evolution of its samples over a fixed prediction horizon. A self‐organizing fuzzy neural network is optimized in its configuration parameters for learning and recognition of these patterns. Four real‐world data sets from three interstate roads are considered for evaluating the performance of the proposed model. A quantitative comparison made with the results obtained by four other relevant prediction models shows a favorable outcome.

A novel wavelet-SVM short-time passenger flow prediction in Beijing subway system

“…The mean absolute percentage error (MAPE), variance of absolute percentage error (VAPE), and root mean square error (RMSE) are chosen to represent the difference between the actual value and the predicted value, as the performance measures for comparisons in the literature [73,19,5].…”

“…Although the multilayer perceptron neural network shows its potential in various problems, elaborate network designs to fit traffic data features are anticipated to upgrade predictive performance [18]. In the later open literature, a lot of improved neural network structures are developed [58][59][60][61][62]15,18,19].…”

“…Lam et al [13] have proven that the Gaussian maximum likelihood model appears to be better than the time series model and nonparametric regression [14] when predicting daily traffic flows in Hong Kong. Then, several algorithms, such as Bayesian networks [15][16][17], neural networks [15,18,19] and support vector machines (SVM) [20][21][22], belong to the artificial intelligence field have been widely applied in the intelligent transportation systems. Furthermore, the related technologies have been successfully used in the field of image processing and 3D model retrieval [23,24].…”

Time series forecasting with a prior wavelet-based denoising step

“…The performance of the proposed model is tested using various time series (the Feigenbaum sequence, a second-order process, a quality control time series) and the model is claimed to outperform the traditional multilayered perceptron and its combination with a statistical denoising criterion. Boto-Giralda et al (2010) use wavelet-based denoising as a part of their algorithm for traffic volume forecasting. The algorithm provides accurate forecasts compared to other benchmark forecasting models.…”

Short-term traffic flow prediction using a self-adaptive two-dimensional forecasting method

“…In general, short-term traffic flow prediction methods can broadly be divided into two groups based on the number of data sources, namely, based on fusing multi-source data, [1][2][3][4][5][6] which try to increase profits to improve forecasting precision, and based on traffic data from a single source data. [7][8][9][10] Recent studies attempt to improve accuracy of prediction methods using multi-source data.…”