Travel Time Forecasting on a Freeway Corridor: a Dynamic Information Fusion Model based on the Random Forests Approach

Details

• Creators:
  Fan, Wei (David) ; Qiu, Bo
• Corporate Creators:
  University of North Carolina at Charlotte. Center for Advanced Multimodal Mobility Solutions and Education
• Corporate Contributors:
  United States. Department of Transportation. University Transportation Centers (UTC) Program
• Subject/TRT Terms:
  • Artificial intelligence
  • Data analysis
  • Data fusion
  • Estimating
  • Freeways
  • Machine learning
  • Probe vehicles
  • Traffic delays
  • Travel time
• Publication/ Report Number:
  2020 Project 01 ; CAMMSE-UNCC-2020-UTC-Project-01
• Resource Type:
  Tech Report
• Geographical Coverage:
  United States
• Edition:
  Final Report
• Corporate Publisher:
  University of North Carolina at Charlotte. Center for Advanced Multimodal Mobility Solutions and Education
• Description:
  Metropolitan areas suffer from frequent road traffic congestion not only during peak hours but also during off-peak periods. Different machine learning methods have been employed in travel time prediction; however, such machine learning methods practically face the problem of overfitting. Tree based ensembles have been applied in various prediction field, and such approaches usually produce high prediction accuracy by aggregating and averaging individual decision trees. The inherent advantages of these approaches are that they can get better prediction results while also having a good bias-variance trade off which can help to avoid the overfitting. However, the reality is that the application of tree-based integration algorithms in traffic prediction is still limited. In order to improve the accuracy and the interpretability of the model, random forest (RF) is used to analyze and model the travel time on freeways. As the traffic conditions often greatly change, the prediction results are often unsatisfactory. In order to improve the accuracy of short-term travel time prediction in the freeway network, a practically feasible and computationally efficient RF prediction method for real-world freeways by using probe traffic data was generated. In addition, the variables' relative importance can also be ranked, which provides an investigation platform to gain a better understanding of how different contributing factors might affect travel time on freeways. This research develops an RF method to predict the freeway travel time by using the probe vehicle-based traffic data and weather data. Detailed information about the input variables and data pre-processing are presented. To measure the effectiveness of proposed travel time prediction algorithms, the mean absolute percentage errors (MAPE) are computed for different observation segments combined with different prediction horizon ranging from 15 to 60 minutes. The parameters of the RF model are estimated by using the training sample set. After the parameter tuning process is completed, the proposed RF model is developed. The features' relative importance shows that the variables travel time 15 minutes before and time of day (TOD) contribute the most to the predicted travel time result. The model performance is also evaluated and compared against extreme gradient boosting method, and the results indicates that the RF always produces more accurate travel time prediction.
• Format:
  PDF
• Funding:
  69A3551747133
• Download URL:
  https://rosap.ntl.bts.gov/view/dot/58260/dot_58260_DS1.pdf
• File Type:
  [PDF - 1.14 MB]
• Collection(s):
  Artificial Intelligence Deployment Center of Excellence (AI COE) ; University Transportation Centers ; US Transportation Collection
• Main Document Checksum:
  urn:sha-512:00da5a3d0c9926613e1bb0e38561b51d57a0bd8c5c54f904d732d0d7273a3a422c5eeac450dc09ccb8a8ea4ee10a1fd12f95f242002aa71263698ca036c74231