Synchronized Entry-Traffic Flow Prediction for Regional Expressway System Based on Multidimensional Tensor

Abstract

Predicting entry-traffic flows synchronously could enable inferences about the changing trends and spatial structure of dynamic traffic flows in an expressway network. This research develops a synchronized entry-traffic flow prediction method for regional expressway systems. The new method first organizes numerous entry-traffic flows as a three-dimensional (time slots, spatial locations, and vehicle types) tensor, then applies tensor decomposition to extract their temporally changing features. After forecasting the temporally changing features, predicted values of entry-traffic flows can be calculated synchronously by tensor reconstruction. Data from hourly entry-traffic flows involving nine vehicle types and 201 spatial locations in a regional expressway system of China are used to discuss the performance of this new method. The results show that the new method could obtain prediction results with high overall accuracy. Comparative experiments indicate that the new method and existing methods (autoregressive integrated moving average, or ARIMA, and Holt-Winters) could generate prediction results with similar accuracy. However, the proposed method has the advantage of reducing the number of time series that need to be handled in the prediction of numerous entry-traffic flows for regional expressway systems. This method might be helpful for administrators to guide and manage vehicles so that they enter the expressway system effectively.

Examining the Effect of COVID-19 on rail freight volume and revenue using the ARIMA forecasting model and assessing the resilience of Indian railways during the pandemic

“…ARIMA Forecasting Model has been extensively used in urban transportation planning by a number of researchers [ 33 , 34 ]. Multiple studies have used ARIMA and other time series forecasting models to predict short-term traffic flow characteristics [ 35 – 37 ], while few researchers have used ARIMA models to predict travel time and travel speed of a road network [ 38 – 41 ]. A number of studies have also used ARIMA Models to predict the freight volume and ridership of public transit modes [ 31 , 32 , 42 – 45 ].…”

Traffic prediction using artificial intelligence: Review of recent advances and emerging opportunities

“…Stakeholders rely on high-level regional insights about the transportation network to plan for future improvements. For example, predicting the traffic flow at entrances to a regional expressway system could be helpful for planners seeking to ensure vehicles enter the system safely and effectively (Gao et al, 2021). Two primary studies attempt to identify the impacts of weather on regional traffic flow concerning region-specific factors (e.g., age, income, road density, hotel density, attraction density) (Ding et al, 2015(Ding et al, , 2017.…”

Travel time prediction for an intelligent transportation system based on a data-driven feature selection method considering temporal correlation

“…According to the literature, numerous studies have been studying prediction of different aspects of traffic such as traffic flow from regional [3][4][5] and network perspectives [6,7], speed [8,9], occupancy [10,11], travel time from path perspective that considers the travel time as the duration that vehicles spend in specific route(s) from entering to exiting [12,13] and origindestination perspective that focuses on vehicles total trip travel time [14,15]. Some other studies have focused on predicting the travel demand [16,17].…”

“…selection in the first OA, 2000 (number of iterations × number of search agents) individual search agents go through a search area of 114 = 11641 different combination to achieve the 12 selected features. This number of combinations is calculated based on the fact that there are five columns in the candidate features' pool, the features in the first columns are always chose, and the potential number of selected features from any of other four column is between 0 and 10.For each one of those search agents from the first OA, another set of 2000 individual search agents go through a search area of 203 = 8000 (three layers with 20 potential neurons in each) different structures to find the optimum structure for ANN with respect to the first OA's agent position. The comparison feature selection methods are also applied to the case study and Fig.9compares their convergence curves to OA2DD's.…”