28 June 2022 in IET Intelligent Transport Systems
Despite complex fluctuations, missing data, and maintenance costs of detectors, traffic volume forecasting at intersections is still a challenge. Moreover, most existing forecasting methods consider an isolated intersection instead of multiple adjacent ones. By accurately forecasting the volume of short‐term traffic, a low‐cost method can be provided to solve the problems of congestion, delay, and breakdown of detectors in the road transport system. This paper outlines a novel hybrid method based on deep learning to estimate short‐term traffic volume at three adjacent intersections. The gated recurrent unit (GRU) and long short‐term memory (LSTM) bilayer network with wavelet transform (WL) noise reduction algorithm (WL+GRU‐LSTM) are used to analyze raw traffic volume data. The WL+GRU‐LSTM is constructed by comparing different machine learning and deep learning methods. A comparative study was used to choose the model’s network structure, training technique, and optimizer type. To prove the model’s accuracy and resilience, it was compared with the leading short‐term traffic forecasting approaches. Experimental results confirm that the WL+GRU‐LSTM model can forecast complex traffic volume fluctuations in different approaches of intersections with an accuracy of over 94%. It also shows better results compared to current methods. The proposed model could replace intermediate loop detectors.
28 February 2017 in Advances in Data Analysis and Classification
30 March 2020 in Transportation Letters
The importance of traffic pattern prediction for traffic management systems has significantly increased in recent years. This paper presents a novel method to find unusual traffic patterns by using topic modeling. We have employed topic models to provide an abstraction of speed camera data from Tehran, the capital of Iran. In this methodology, topic modeling is applied to days of weeks and months in a year and extracts weekly and monthly traffic patterns. Analysis of the abstract descriptions and their adaptation to actual urban traffic patterns prove the effectiveness of the proposed method. The model training convergence is also practically verified. Based on our experiments, our method achieves an accuracy of 99% in detecting abnormal conditions, which indicates the fitness of the topic modeling abstraction. Such a powerful abstraction capability can be exploited as a method for data comparison and search procedures.
1 June 2024 in Journal of Transportation Engineering, Part A: Systems
16 September 2024 in Information Sciences
30 October 2014 in Machine Vision and Applications