A Machine Learning-Based Overlay Technique for Improving the Mechanism of Road Traffic Prediction Using Global Positioning System
Global Positioning System (GPS)-based road traffic prediction is one of the predominating technology in the modern technological era, which facilitates smooth navigation and reduces mobility time. Google Maps is used worldwide for traffic congestion and delay prediction which relies upon the GPS location of the individual’s smartphone to predict traffic congestion and delay by stored data and current GPS locations. However, this method sometimes malfunctions due to the uneven distribution of passengers in different vehicle types on the roadway as there are far more passengers in buses as compared with trucks, if few buses are present in the traffic stream then it will show congestion and delay in traffic. So, it is hard to correctly predict the congestion and delay in traffic without using classified vehicle count as the ratio of the area occupied by the vehicle on the roadway and the number of passengers in it is unevenly distributed for different vehicle types. Google Maps have some limitations as it does not incorporate details regarding the classified vehicle count and categories of vehicles as there are distinct categories of vehicles operating on the roadways, with varying sizes, speeds, and passenger capacities. Thus, it would be beneficial to overlay the information of GPS localization, using Google Maps, with classified vehicle count and vehicle categories to estimate better road traffic congestion and delay. Thus the augmentation of Google Maps is required by integrating the classified traffic volume count with categories of vehicles, the present work envisages the same. For the present study, two mid-sized Indian cities in the state of Uttar Pradesh (Varanasi and Gorakhpur) were selected due to the diverse nature of mixed road traffic. For classified vehicle count data, video recording was carried out by using video recording cameras at various sites in both cities. The data of classified vehicles for both directions of traffic streams were manually counted by project staff from the video recordings and GPS coordinates were also integrated with datasets. Subsequently, various other hand-crafted features were extracted before training the machine learning-based forecasting models (ARIMA and SVM) for traffic volume prediction for a specified GPS location. The classified road traffic vehicle count was predicted using previously observed values, thereby helping in making a good decision about route selection and traffic management. Further, this work annotates the forecasted data overlay with GPS value as per the traffic condition to build a XGBoost-based classification model. The build classifier can classify the road conditions in real-time. The rigorous experimental results and real-world evaluation depicted the effectiveness of the proposed technique on the collected dataset
Year of publication: |
[2023]
|
---|---|
Authors: | Pandey, Amar Deep ; Kumar, Brind ; Parida, Manoranjan ; Chouksey, Ashish Kumar ; Mishra, Rahul |
Publisher: |
[S.l.] : SSRN |
Saved in:
freely available
Saved in favorites
Similar items by person
-
Pandey, Amar Deep, (2023)
-
Connecting South and Southeast Asia: Implementation challenges and coordination arrangements
Thuzar, Moe, (2014)
-
BCIM-economic corridor : the road ahead
Bhatia, Rajiv K., (2015)
- More ...