A Comparative Analysis of Metaheuristic Approaches for Sensors Deployment Problem on Transport Networks

The need for traffic flow data is essential for proper traffic management and control. Travel time estimation and early response to possible traffic incidents can be achieved with deployment of appropriate number of detectors, and placing them on optimal locations on traffic network. With more detectors located the level of accuracy of the data obtained increases, while at the same time requires more investment and maintenance costs. The detectors should be deployed in such a way to appropriately sample traffic conditions, and also provide travel time estimation with the lowest possible error. On the other hand, traffic authorities have a tendency to reduce the number of detectors located on the network in order to achieve investment savings. The proposed model provides the most suitable detector locations on a road section, while minimizing travel time estimation error with limited available funds which are considered in the model constraints.The Bee Colony Optimization metaheuristic was used to solve the sensor deployment problem, and its variant based on solution improvement, BCOi. The results obtained using BCOi metaheuristics were compared with the results obtained using the Simulated Annealing (SA) metaheuristics. In terms of the CPU time, BCOi outperformed the SA algorithm, while in comparable operating time the BCOi algorithm achieved better solutions to larger scale problems. The applications of both algorithms were tested on real case study data on a section of the E-763 road in the Republic of Serbia.

Language

  • English

Media Info

Subject/Index Terms

Filing Info

  • Accession Number: 01777257
  • Record Type: Publication
  • Files: TRIS
  • Created Date: Jul 23 2021 3:25PM