Near‐Term Travel Speed Prediction Utilizing Hilbert–Huang Transform

Abstract

Accurate short-term prediction of travel speed as a proxy for time is central to many Intelligent Transportation Systems, especially for Advanced Traveler Information Systems and Advanced Traffic Management Systems. In this study, we propose an innovative methodology for such prediction. Because of the inherently direct derivation of travel time from speed data, the study was limited to the use of speed only as a single predictor. The proposed method is a hybrid one that combines the use of the empirical mode decomposition (EMD) and a multilayer feedforward neural network with backpropagation. The EMD is the key part of the Hilbert-Huang transform, which is a newly developed method at NASA for the analysis of nonstationary, nonlinear time series. The rationale for using the EMD is that because of the highly nonlinear and nonstationary nature of link speed series, by decomposing the time series into its basic components, more accurate forecasts would be obtained. We demonstrated the effectiveness of the proposed method by applying it to real-life loop detector data obtained from I-66 in Fairfax, Virginia. ous forecasting techniques. Rigorous testing of the distribution of prediction errors revealed that the model produced unbiased predictions of speeds. The superiority of the proposed model was also verified during peak periods, midday, and night. In general, the method was accurate, computationally efficient, easy to implement in a field environment, and applicable to forecasting other traffic parameters.

Stochastic Approach for Short-Term Freeway Traffic Prediction During Peak Periods

“…Stathopoulous et al [34] employed a fuzzy rule-based system to nonlinearly combine traffic flow forecast results from an online adaptive Kalman filter and an artificial NN model. Given the highly nonlinear and nonstationary nature of link speed series, Hamad et al [12] proposed a hybrid method combining the use of the empirical mode decomposition and a multilayer feedforward NN with BP for short-term travel speed prediction. Boto-Giralda et al [3] developed a wavelet-based denoising self-organizing NN model for traffic volume time-series forecasting.…”

Exploring time variants for short-term passenger flow

“…From the experimental results, it was revealed that the hybrid approach, which takes advantage of EMD, outperforms the traditional traffic volume prediction model involving a simulation model and time series method. However, Hamad et al (2009) have only focused on building a forecasting model of short-term traffic volume. The time variants of passenger flow, which plays an essential role in transportation systems, still remains insufficiently investigated.…”

“…For example, references (Huang et al, 2003a;Wu, 2007;Guhathakurta et al, 2008; have utilized EMD to analyze financial time series. Additionally, Hamad et al (2009) has applied EMD to analyze traffic volume data. Hamad et al (2009) applied a combined approach of EMD and back-propagation neural networks to predict traffic volume by using a set of real-life loop detector data.…”

“…Additionally, Hamad et al (2009) has applied EMD to analyze traffic volume data. Hamad et al (2009) applied a combined approach of EMD and back-propagation neural networks to predict traffic volume by using a set of real-life loop detector data. From the experimental results, it was revealed that the hybrid approach, which takes advantage of EMD, outperforms the traditional traffic volume prediction model involving a simulation model and time series method.…”

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

“…Among these methods, neural network models stand out mainly for their ability to describe the indeterministic and complex nonlinearity of traffic flow time series (Dharia and Adeli, 2003; Jiang and Adeli, 2005). Neural network applications to short‐term traffic flow forecasting extend from one single network structure using techniques such as backpropagation (Dochy et al, 1995; Dougherty and Cobbett, 1997) or radial basis function (Park et al, 1998; Chen and Grant‐Muller, 2001) to more complex structures that include diverse methodologies such as autoregressive models (Voor et al, 1996), fuzzy theory (Yin et al, 2002; Park, 2002), wavelet functions (Jiang and Adeli, 2005; Xie and Zang, 2006), online Kalman filter (Stathopoulos et al, 2008), k‐means algorithm (Vlahogianni et al, 2008), Hilbert‐Huang transform (Hamad et al, 2009), or a combination of various neural network models (Lee et al, 2004; Zheng et al, 2006). In most of the aforementioned models, input data come mainly from loop detectors placed at the same road link locations where predictions are made.…”