Traffic-flow prediction via granular computing and stacked autoencoder
- PDF / 1,554,273 Bytes
- 11 Pages / 595.276 x 790.866 pts Page_size
- 76 Downloads / 186 Views
ORIGINAL PAPER
Traffic‑flow prediction via granular computing and stacked autoencoder Jianhua Chen1,2 · Wenjing Yuan1,2 · Jingjing Cao1,2 · Haili Lv1,2 Received: 15 December 2018 / Accepted: 12 April 2019 © Springer Nature Switzerland AG 2019
Abstract Accurate traffic-flow prediction is essential to traffic management. Traffic data collected in very short intervals normally contain high variability, while common preprocessing approaches applied within a window such as simple average or median operator are unable to obtain sufficient latent information from original data. Moreover, the prediction performance of shallow neural network is not satisfying, since its capacity to represent the temporal–spatial correlation of mass traffic data is insufficient, and its adaptation capacity to noisy data is relatively poor. In this paper, fuzzy information granulation (FIG) and deep neural network are combined to solve these two issues. To be specific, FIG is utilized to process original data series and extract granules, which denote the trend and fluctuation range of each time window. Then, stacked autoencoder is combined to obtain the predictive results based on processed granules, especially, a multi-output mechanism is designed to predict all granulation parameters simultaneously, which makes better use of the correlation of diverse inputs. A real-world traffic volume data set is applied to conduct an empirical study, and the experimental results illustrate that based on the proposed method, the interval prediction of the traffic-flow fluctuation range is realized, and superior traffic trend prediction performance is achieved. Keywords Traffic-flow prediction · Deep learning · Stacked autoencoder · Fuzzy information granulation · Granular computing
1 Introduction As the core technology of Intelligent Transportation Systems (ITS), short-term traffic state forecasting is very important for efficient transportation management (Ma et al. 2016). To some extent, accurate prediction of traffic flow can alleviate road congestion and help to establish a promising real-time traffic assignment system, which is essential to the efficient * Jingjing Cao [email protected] Jianhua Chen [email protected] Wenjing Yuan [email protected] Haili Lv [email protected] 1
School of Logistics Engineering, Wuhan University of Technology, Wuhan, China
Engineering Research Center of Port Logistics Technology and Equipment, Ministry of Education, Wuhan, China
2
guiding of traffic flow. However, traffic flow is characterized by high uncertainty and nonlinearity, and short-term trafficflow prediction is a more complex issue, which denotes forecasting the future flow based on historical data series in short time period, which is normally shorter than 15 min (Smith and Demetsky 2002). Traffic-flow parameters to be predicted include vehicle velocity, road occupancy, and vehicle volume in a specific section of highway; since the traffic volume data are relatively visualized and easy to be applied into traffic management assistance,
Data Loading...
